Segment 2 Of 2 Previous Hearing Segment(1)
SPEAKERS CONTENTS INSERTS
Page 7 PREV PAGE TOP OF DOC Segment 2 Of 2
U.S. SUPERCOMPUTER EXPORT CONTROL POLICY
House of Representatives,
Committee on National Security,
Washington, DC, Thursday, November 13, 1997.
The committee met, pursuant to call, at 10:08 a.m. in room 2118, Rayburn House Office Building, Hon. Floyd Spence (chairman of the committee) presiding.
OPENING STATEMENT OF HON. FLOYD D. SPENCE, A REPRESENTATIVE FROM SOUTH CAROLINA, CHAIRMAN, COMMITTEE ON NATIONAL SECURITY
The CHAIRMAN. The committee will please be in order. Good morning.
On October 28, the Director of OMB sent me a letter outlining the administration's concerns with H.R. 1119, the National Defense Authorization Act for fiscal year 1998. In addition to the veto threat on the depot maintenance issue, the letter also cited the administration's strong objections to the conference compromise on the issue of supercomputer export controls. Interestingly enough, over the course of several months of communication with the administration concerning the defense authorization conference, supercomputers were not raised as a significant issue until then.
Page 8 PREV PAGE TOP OF DOC Segment 2 Of 2
Therefore, we are holding this hearing this morning to attempt to gain a better understanding of why the administration continues to object to what is modest and reasonable legislation to address a serious problem.
Since the administration relaxed its policy last year on supercomputer exports, there have been numerous revelations about the unauthorized shipment or diversion of U.S.-made supercomputers to countries and entities of proliferation concern. We have learned that United States supercomputers have been inappropriately shipped to military research facilities in China and nuclear weapons laboratories in Russia. By the admission of Russian officials, these computers will be used to help maintain Russia's nuclear weapons stockpile.
Most recent press stories also indicate that an additional 16 United States-made high-performance computers were illegally obtained by a Russian nuclear weapons laboratory using European middlemen in violation of United States export control regulations.
The true impact of these transfers on the ability of other countries to develop weapons that pose a threat to U.S. interests may never be fully known. However, it seems to me that it is in our national interest to find out. In this regard, I am deeply concerned over the reports that the Commerce Department has refused to provide all relevant information to the intelligence community in order to help it assess the potential military consequences of these transfers. I hope these reports are false.
I firmly believe that these unauthorized transfers have been facilitated, if not encouraged, by the administration's own relaxation of supercomputer export controls. As a result, earlier this year, Mr. Dellums and I jointly sponsored legislation designed to help close the loophole in the current export control process that allowed these transfers to occur. The legislation, which has since been modified in conference with the Senate, passed the House earlier this year with an overwhelming vote of 332 to 88. Yet the administration apparently objects even to the compromise.
Page 9 PREV PAGE TOP OF DOC Segment 2 Of 2
Our legislation was designed to correct the most glaring flaw in the administration's policy, which is the honor system approach that places industry in the inappropriate role of determining for themselves the true identity of supercomputer customers in countries of proliferation concern. This arrangement has led, at a minimum, to dozens of U.S. supercomputers being sold to users of questionable reputation, transactions that under the administration's policy take place with little or no visibility on the part of the U.S. Government.
As incomprehensible as it may seem, under this relaxed policy, the administration did not know that a United States-manufactured supercomputer had been exported to one of Russia's premier nuclear weapons laboratories until the Russian Minister of Atomic Energy revealed it during a press conference. Until this committee and others in Congress began asking obvious questions, the administration was unsure of how many supercomputers had been exported to China and apparently did not know that at least one had been diverted to a Chinese military institute. Given this pattern, I suspect it will be years before we have a full appreciation of how many supercomputers may have slipped through the loopholes that characterize the administration's relaxed export policy.
Some of these incidents are now the subject of Justice Department criminal investigations. The rest will require the watchful eye of our intelligence community once it has received all pertinent information.
Contrary to the inflated rhetoric associated with opponents of this provision, the legislation merely requires that supercomputer exports to countries of proliferation concern, so-called tier III countries, first go through a 10-day review period by the Federal Government. Such a short review would still permit our national security, arms control and diplomatic agencies the opportunity to assess whether an otherwise innocent-looking transaction poses security concerns. If such a concern is raised during the review, then the normal export license process would apply. However, the important point that bears repeating is that nowhere in this legislation are supercomputer exports to any countries banned or prohibited. All the legislation does is require that the Federal Government pay attention to and review the flow of this sensitive technology to these nations identified as a proliferation concern.
Page 10 PREV PAGE TOP OF DOC Segment 2 Of 2
Striking a balance between protecting national security interests and promoting U.S. exports has always been contentious. Given the commanding share of the global information technology market that U.S. companies currently hold, we should not unnecessarily inhibit or damage American trade competitiveness. However, I am firmly convinced that the legislation in question has struck an appropriate balance by granting the President sufficient flexibility to reflect the rapid advance of technology as well as any future changes in the status of so-called tier III nations. Given the stakes involved, asking industry to sacrifice 10 days to ensure that the Federal Government has an opportunity to at least review supercomputer exports does not strike me as being too high a price to pay.
To address these issues today the committee will hear from two panels. Our first panel consists of a Department of Defense and a Department of Commerce witness, who will present the administration's views on the computer issue.
Mr. William Reinsch, Under Secretary for Export Administration, Department of Commerce; and Dr. Mitchel Wallerstein, Deputy Assistant Secretary for Counterproliferation Policy of the U.S. Department of Defense.
Our second panel will consist of outside witnesses with extensive knowledge in the fields of nonproliferation and export control policy. Mr. Gary Milhollin, director of the Wisconsin Project on Nuclear Arms Control, and Mr. Stephen Bryen, president of Delta Tech and former Director of the Defense Technology Security Administration.
Before proceeding, I would like to first recognize the distinguished ranking member, the gentleman from California, Mr. Dellums, for any remarks he would like to make.
Page 11 PREV PAGE TOP OF DOC Segment 2 Of 2
STATEMENT OF HON. RONALD V. DELLUMS, A REPRESENTATIVE FROM CALIFORNIA, RANKING MEMBER, COMMITTEE ON NATIONAL SECURITY
Mr. DELLUMS. Thank you very much, Mr. Chairman. Forgive my hoarseness.
Mr. Chairman, I wish to join with you in welcoming our distinguished witnesses today. I would like to welcome each of them back to the committee and I look forward to their continued illumination of the issue that is before us. It is obvious to all that the committee remains vitally interested to learn as much as we can concerning the report of sales of additional U.S.-origin high-performance computers to countries of proliferation concern.
While I remain convinced that the reports of such sales received earlier this year warranted some fixes to the administration's current export control policy, I believe these more recent reports highlight the prudence of having moved forward with such fixes.
I appreciate, Mr. Chairman and members of the committee, the good working relationship that my office had with the administration over the past half-year concerning our differences of opinion on the scope and scale of these fixes, and I compliment the administration on its sincere efforts to meet our concerns by the publication of some end users of proliferation concern, integrated interaction with industry to highlight the dangers of exports to such destinations. I remain, Mr. Chairman, of the view that these efforts, however well intend ed, did not sufficiently close the gap, and I believe that our legislative efforts are required to achieve that goal.
Page 12 PREV PAGE TOP OF DOC Segment 2 Of 2
While reasonable people can disagree with the outcome of the defense authorization conference report recently adopted in both the House and the Senate, I believe it constitutes an acceptable balance of flexibility and control on the important issue of controlling the export of this technology. As members of this committee will know, the compromise outcome left standing the requirement that high-performance computer sales to countries of proliferation concern receive at least a cursory Government review before export. And you, in your opening remarks, Mr. Chairman, alluded to that in specific terms.
The compromise provides flexibility both for the President to elevate the threshold for computing power that would be controlled and to remove countries from the so-called tier III list. This flexibility would allow the President to take into account changes in availability of technology in order not to unjustly restrict the economic opportunities of U.S. firms. It would also allow the President to conform our Nation's policy to ensure that it targets nations of proliferation concern.
The administration has communicated its disagreement with the conference outcome, not surprising in the light of its continued view of the efficacy of its implemented policy as modified this summer. The administration has the right to hold its views, and I believe it is noteworthy that in raising its oral and written objections, the administration representatives have never, to this gentleman's knowledge, articulated a vetoed threat in direct relationship to the matter before this committee.
Mr. Chairman, by this legislative effort we have altered the terrain upon which the issues before us can be assessed. It is important for us to know what has transpired and what the administration's plans are with respect to the particular transaction. What more can be done, though, it seems to me, has been anticipated and answered in the legislation that we have put before the President in the form of the fiscal year 1998 defense authorization bill.
Page 13 PREV PAGE TOP OF DOC Segment 2 Of 2
In that context, I look forward to hearing from our distinguished witnesses today on this topic of continued interest, significance and import.
And with those remarks, Mr. Chairman, I would yield back the balance of my time.
The CHAIRMAN. Thank you very much, and without objection, the prepared statements of the witnesses will be presented for the record; and, Mr. Reinsch, you can proceed as you like.
I understand you have a little trouble with your voice?
STATEMENTS OF HON. WILLIAM REINSCH, UNDER SECRETARY FOR EXPORT ADMINISTRATION, ACCOMPANIED BY AMANDA DEBUSK, ASSISTANT SECRETARY FOR EXPORT ENFORCEMENT, DEPARTMENT OF COMMERCE
Secretary REINSCH. Yes, as you can tell, Mr. Chairman, I am not in the best of shape this morning. I have a cold. If you do not mind, I would like our newly confirmed Assistant Secretary for Enforcement, Amanda DeBusk, to deliver my oral statement, and then I will be happy to answer questions, shorter answers than usual, I think.
The CHAIRMAN. Yes, sir, that will be fine.
Secretary DEBUSK. Thank you, Mr. Chairman. I appreciate the opportunity to discuss once again export controls on high-performance computers.
Page 14 PREV PAGE TOP OF DOC Segment 2 Of 2
The Bush administration controlled computer exports to most destinations at the level of then commonly available personal computers. A supercomputer was defined as performing at more than 195 MTOPS. Today many, if not most PC's, exceed that level. When President Clinton took office, congressional leaders of both parties urged reform of this clearly outdated standard.
I would like at this point to insert in the record a letter to the President from congressional leaders. Thank you.
[The information referred to can be found in the appendix on page 70.]
Secretary REINSCH. An initial reform in February 1994 raised the control level to most destinations to 500 MTOPS and the definition of supercomputer to 1,500 MTOPS. But these thresholds were soon overtaken by advances in computer technology.
In January 1996, the administration again adjusted controls to reflect availability and national security relevance. This is the structure the committee is familiar with: MTOPS thresholds of 2,000, 7,000, and 10,000, depending on the country tier.
Numerous safeguards underpin this policy. A license is required for the export of a computer of any level, regardless of destination, if the exporter knows it will be used in a proliferation-related activity. Last year we established a process for publishing lists of end users involved in the development of weapons of mass destruction. We have been publishing the names of such entities since February and expect to continue doing so.
Page 15 PREV PAGE TOP OF DOC Segment 2 Of 2
Finally, we have established a recordkeeping requirement specific to exports of high-performance computers. It compels maintenance of export transaction details and has a comprehensive outreach program. Through that program, Bureau of Export Administration [BXA] personnel meet with industry representatives, alert industry to special problems, seek their cooperation, help them know their customer, educate them on the legal requirements, and share information.
Use of these safeguards has been rigorous. We maintain a watch list against which every export license application is screened to clear all end uses and users, and we conduct prelicense checks and postshipment verifications when appropriate.
Following reports of certain tier III countries' acquisition of HPC's for end users that would, in our judgment, require an individual license, we have further stepped up enforcement efforts. As part of this process, we are analyzing transaction details from all U.S. companies identified as producers of HPC's.
The producers reported exporting 1,437 systems, totaling $785 million between January 1996 and March 1997. Of those, 91 were exported to tier III countries. Of those 91, 47 went to China and 10 went to Russia. These are the statistics I have shared with the committee previously.
BXA has opened four investigations on HPC exports, two involving China and two involving Russia. All these investigations are in the hands of the Department of Justice. China has cooperated with us and the Sun Microsystems computer was returned to the United States on November 9. The Russian officials have indicated that Russia is considering moving the computers in question to facilities where unclassified research is being done. We hope to continue discussions regarding the details in the near future.
Page 16 PREV PAGE TOP OF DOC Segment 2 Of 2
Thus far in 1997, BXA has requested 22 postshipment verifications of HPC's exported to tier III countries. Seventeen have been completed, all with favorable results. The remainder of the requests are pending at our embassies.
Our previous work has suggested that it is impossible to control the spread of computers below the level of 4,000 MTOPS and that, in any case, below 7,000 MTOPS the speed of the computer makes little difference in weapons design. To catch computers used in weapons design, we would have to set levels that control the PC's you can buy at your corner store. But since they are also made by the thousands in Taiwan, Korea, and China, even that unilateral step would be ineffective.
In 1995, we determined that at that time the lower boundary of controllability was between 4,000 and 5,000 MTOPS, rising to 7,000 this year. The Wassenaar arrangement members have agreed to decontrol all computers up to 2,000 MTOPS and to require reporting only for computers above 4,000 MTOPS.
Those levels become even more questionable when we look at technological developments. Chips the size of a dime are now capable of operating at 1,300 MTOPS and the industry is building and will begin to market next year a single chip capable of 2,000 MTOPS. At least one company plans to offer laptop computers capable of 2,000 MTOPS within the year. I would not be surprised within a year or two to find that the 2,000 MTOPS level catches high-performance video games, which need quick processing times for their advanced graphics.
Thus, as I have said before, our conclusion is that computer controls must be related to reality. However much we would prefer to have it otherwise, we must not delude ourselves; we cannot control the uncontrollable. The widespread availability of microprocessor and computer capabilities, compounded by computing power aggregation technologies, such as networking, clustering and parallel processing, make attempts to control at increasingly high levels fruitless. Trying to do so undermines the credibility of our export control regime and diverts scarce enforcement resources that could be better spent on matters more directly related to our national security.
Page 17 PREV PAGE TOP OF DOC Segment 2 Of 2
Finally, Mr. Chairman, I would be remiss if I did not take this opportunity to restate the administration's opposition to the computer language contained in the Defense authorization. It sets forth a control regime which is unrealistic both philosophically and procedurally. Let me be a bit more specific about the congressional language.
First, the provisions infringe significantly on the President's ability to conduct foreign policy. While it ostensibly gives the President flexibility to change the list of countries in tier III, in fact, Russia, China, India, Israel, and Pakistan could not be removed.
While we have no plans to move any of these countries from tier III, this provision will seriously damage our relations with them. Locking them statutorily into subordinate status precludes the President from moving them to a different tier regardless of any changes in their behavior. It will also make it harder for us to persuade them to join multilateral regimes.
Second, the waiting periods in the bill are an inappropriate and unnecessary restriction on Presidential flexibility and make no technological sense. The waiting period between congressional notification of a change in the MTOPS level for tier III countries and its implementation is 180 days. In an industry where product life can be 18 months to 2 years, a 6-month delay and the uncertainty that comes with it makes no technological sense and would seriously handicap industry's ability to market its products. This lengthy period is in stark contrast to the 30 days required for arms sales or changes in the U.S. munitions list. Notification for arms sales is limited to those over $14 million.
Almost all HPC's are at the levels we are talking about, and they would cost less than that. It is hard to understand why the Congress would require 30 days notification for a change in controls on an M–1 tank or an F–15, but 180 days for a computer.
Page 18 PREV PAGE TOP OF DOC Segment 2 Of 2
Third, the requirement to conduct postshipment checks will become an extraordinary resource burden, is unadministrable and unnecessary. The provision requires postshipment checks on all computers over 2,000 MTOPS going to tier III countries regardless of whether that level is increased in the future. Given the rate of growth of computer sales, the size of tier III covering 50 countries and the prohibition on removing the biggest countries from it, this will become an unadministrable requirement within a year. It is not helpful to either Congress or the executive branch to require reporting on activities that cannot fully be carried out.
In addition, there are many circumstances where it is a waste of time and money to conduct a check. The most obvious is when the exporter performs regular service and maintenance on a computer and is in a position to know on a regular basis whether the computer is properly located. The executive branch should have discretion not to perform checks when they are unnecessary.
Mr. Chairman, these provisions are an attempt to micromanage export policy that will harm our own computer industry far more than it will block tier III countries' access to computers. The United States dominates the high-performance computer industry today, in part due to our export control reforms.
A strong computer industry has many benefits for the United States. We lead in the information technology sector and our own national security benefits from a robust HPC industry which can meet our defense needs with lower costs, quicker delivery, and more alternatives than would otherwise be the case.
Page 19 PREV PAGE TOP OF DOC Segment 2 Of 2
We have allowed our industry to fill the legitimate civil demand that exists for these machines by dominating the civil market. We have made it expensive and unattractive for other countries to build their own HPC industries. Given the global spread of technology and widespread microprocessor manufacturing capability, I expect that, if enacted, this provision will be an indirect subsidy for Japan's HPC industry, and we can expect to see stronger HPC industries in Russia, China, and India.
It was the administration's intent to set control levels and licensing conditions consistent with national security and proliferation risks, and with an understanding of the limits of what can be expected of export controls. With the passage of time and rapid advances in technology, we must regularly reassess our policy to ensure that it continues to meet those objectives. Congress' action in this case would make that process more difficult to the detriment of one of our critical industries.
That completes my statement.
The CHAIRMAN. Thank you.
[The prepared statement of Secretary Reinsch can be found in the appendix on page 45.]
The CHAIRMAN. Mr. Wallerstein.
STATEMENT OF MITCHEL B. WALLERSTEIN, DEPUTY ASSISTANT SECRETARY FOR COUNTERPROLIFERATION POLICY, DEPARTMENT OF DEFENSE
Page 20 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. WALLERSTEIN. Thank you, Mr. Chairman, for the opportunity to discuss DOD's view on high-performance computer export controls.
The Defense Department views export controls as a critical part of our national security and defense policy. It is important that we deny potential adversaries those technologies and items that would provide them with key military capabilities that would enable them to seriously threaten U.S. forces deployed abroad and, of course, our national security more generally. DOD has always given serious attention to export controls on high-performance computers because their capabilities are of significance in applications ranging from acoustic submarine detection to missile defense.
We also recognize, however, that computer technology continues to advance very rapidly, and ever-increasing computing capabilities are widely available on a global basis. Computer users no longer have to rely on large, expensive, single-vector supercomputers to solve complex problems. Rather, advances in microchip processing power have given computer manufacturers the ability to offer flexible, adaptable, and powerful workstations to perform high-performance computing applications.
Thus, we must recognize that export controls only advance U.S. interests if they can be effective. Ineffective export controls only create an illusion of security protection where, like a Maginot Line, none exists and they waste scarce resources of the Government and of industry in trying to enforce export controls without compensating national security benefit.
Page 21 PREV PAGE TOP OF DOC Segment 2 Of 2
Before we last revised computer controls in 1995, the Department of Defense conducted an unprecedented review of over 600 DOD military programs to determine how the Department actually uses computers and where our critical performance thresholds actually lie. The categories of applications investigated included surveillance, target detection and recognition, submarine design, aerodynamic vehicle design, explosives research, weapons guidance, and cryptology.
Our findings indicated that there were essentially two clusters of critical DOD applications, one around 10,000 million theoretical operations per second what is referred to as MTOPS, and the other around 20,000 MTOPS. We also found that a significant number of applications below 10,000 MTOPS are increasingly being performed in DOD on workstations, commercially available workstations.
As part of a larger study by the administration, which included all of the departments and agencies with national security, nonproliferation, and export control responsibilities, we also looked at trends in computer technology. We made judgments on the effectiveness of controls by taking the following security and market factors into account: First, the extent of worldwide production and number of computer units in the field, as well as their size, their portability, and their cost; second, the ease of upgrades to higher performance levels without manufacturing support; third, availability of commodity high-performance communications links to network together individual computers; and fourth, the increasing parallelization of important national security applications.
We concluded at that time that workstations up to 7,000 MTOPS were becoming widely available, and therefore uncontrollable, from a variety of sources, and that this level of computing performance could be obtained relatively easily worldwide by a variety of means—for example, by buying lower-level workstations and upgrading them, or by buying the components and piecing together a high-performance computer.
Page 22 PREV PAGE TOP OF DOC Segment 2 Of 2
For example, a low-level workstation, two to four processors, that are readily available through commercial distribution networks globally can be easily upgraded with plug-in words that are widely available commercially and, therefore, uncontrolled to create a system with 64 or more processors that are capable of achieving performance levels in the 7,000 MTOPS range. Numerous foreign manufacturers can purchase microprocessors and other components and build relatively sophisticated machines.
Microprocessor performance has more than quadrupled in the last 5 years, from 80 MTOPS in 1993 to over 450 MTOPS today. And new developments in high-speed communications, for example, high-performance parallel interconnect, HIPPI switches; and fiber distributed data interconnect, the FDDI switches, provide the means to connect many low-level workstations to achieve performance levels above 7,000 MTOPS. In fact, many DOD research labs are now using such commercially available workstations and interconnect technology to develop and test weapons systems.
Our findings on national security applications and technology trends were the basis for our consensus U.S. Government decision to establish the export control levels on high-performance computers that we have in place today. The Department continues to fully support these control levels because we believe that they are technically realistic and that they also meet our national security and our nonproliferation objectives.
Specifically, the control methodology is designed to serve two purposes:
First, it permits the Government to calibrate control levels and licensing conditions to the national security or proliferation risks posed by specific destinations or end users. The most stringent controls and licensing conditions are focused, of course, on countries that pose significant proliferation or other security risks. We have the ability to deny access to U.S.-origin computing power needed for critical military applications to such countries.
Page 23 PREV PAGE TOP OF DOC Segment 2 Of 2
I should note, by the way, that we also continue to maintain our controls under the enhanced proliferation control initiative [EPCI] under which computer exporters who know or have reason to know that they are selling to end users involved in proliferation projects must obtain a license for any level of computer.
Our second purposes is to enhance U.S. national security and to preserve the U.S. computer industrial base by ensuring that legitimate computer exports are not impeded.
Finally, Mr. Chairman, I would like to comment briefly on the legislative language on high-performance computer export controls in the Defense authorization conference bill. I particularly question the advisability of the 180-day congressional notification period for any export control policy changes in such a fast-moving technology sector as computers.
I am a firm supporter of appropriate congressional consultation on matters of national security. Nevertheless, I would hope that a more reasonable time period of 30 days would suit both our requirements as well as those of the administration, because that would provide an opportunity for both appropriate congressional oversight and for the administration to do its work in making changes in control parameters as they become necessary to keep pace with rapidly advancing technologies.
I want to underscore the continued commitment of the Department of Defense to strong and effective export controls. However, the approach taken in the defense authorization conference bill would significantly impair the President's flexibility, and indeed that of the Department of Defense, to ensure that export control policies and procedures are implemented in a manner that protects our national security interests without damaging the continued economic viability of the U.S. computer industry, which is of course a vital element of the U.S. defense industrial base. We remain committed to working with the Congress to address these issues in a manner that does maintain the necessary flexibility that we need on these important national interests.
Page 24 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. Chairman, this concludes my formal statement. I am happy to respond to your questions.
[The prepared statement of Mr. Wallerstein can be found in the appendix on page 50.]
The CHAIRMAN. Mr. Dellums.
Mr. DELLUMS. Thank you, Mr. Chairman. First, let me concede one point, that we do live in a tiny world and we do live in a global world, because one of the arguments is, you can obtain 20,000 MTOPS here, 10,000 MTOPS here, et cetera, et cetera.
We live in a small world. We understand that, and I am prepared to concede that point. I would hasten to add, however, if we are the only superpower standing and we assume that we have a leadership role in the world, how then do we say to other nations in the world, you need to control your exports of these technologies because of the potential proliferation concern if we do not show them that we have export control?
So I would suggest to you that the answer to that question is that we play an important role in the world and we ought to be leaders and we have a responsibility to step up to that.
The President of the United States, the Secretary of State, we have other fora within which we can sit down with other nations to say, look, in your country you have got this level of capability out there. We are trying to control ours. We suggest that you attempt to control yours because this world is a very tiny place. Because when I listen to the testimony, the assumption is, this is a great big world and we do not communicate. But I think we do and we have to show some leadership.
Page 25 PREV PAGE TOP OF DOC Segment 2 Of 2
Second point. In your opening remarks, Mr. Reinsch, you made the comment, and I wrote it down here, we cannot control the uncontrollable. Before I go further, I would like you to amplify upon that, because if you are saying that industry has created such an incredible level of technology, moving at such an extraordinary speed that our antiquated governmental processes cannot get out in front of it, then that is a powerful statement and we need to try to understand if that is exactly what you are saying.
My approach to this whole thing is, that may be correct, but I am not prepared at this point to lay down on that altar and say technology is so extraordinary that we as human beings cannot figure out even in our slow-moving political process how to gain some control. That is a concession that is extraordinary with massive implications.
So I would like to go back to your comment, we cannot control the uncontrollable, to determine if what you are really saying is that there is no way on this Earth that we can control export of high performance computers to tier III countries. Civilian end use. Civilian end user. Because everything else you already have to have control.
Mr. REINSCH. I would not take the statement quite as far as you did, Mr. Chairman. We would not argue that high performance computers at any level are uncontrollable. We believe that they are controllable at the levels that we have set and we continue to try to match control levels to what we believe is viable.
The two issues that are complicating our life in this regard—and these are developments that were under way during the President's 1995 decisionmaking process, but developments that will be even more obvious when we review this issue again—are the explosion and advances in microprocessor technology and in scaleability technologies.
Page 26 PREV PAGE TOP OF DOC Segment 2 Of 2
The problem from the standpoint of high performance computers is really not in the box, the computer box, but in the chip. And when you have chips that are getting smaller yet—the line widths are getting narrower, they have more and more information capacity on them—you are developing smaller and smaller units with higher and higher performance levels.
Chips by and large are not controlled, and have not been controlled since 1990. Scaleability is the term that underlies a number of the comments that both our statements and, I believe, the chairman's statement made. That is the capacity to take a computer and expand its performance to a higher level through either an additional memory, additional boards, additional processors, or by linking it with other computers. And there has been an explosion in the technology and the software also that permits that to happen.
So what we are faced with is not only companies building machines at a higher performance level but an increasingly widespread capacity throughout the world to take machines at the PC level or at the 1,500 MTOPS workstation level, for example, and link them together and produce, ultimately, computing capability that in the aggregate is larger than our licensing requirements.
Now, our licensing requirements catch those situations. If you are going to do something to a machine that would enhance its capability beyond a level where a license would be required, you have to get a license for that enhancement. But as you can also see, that—particularly since those enhancements are often not made here, but are made in other parts of the world, that is a more difficult part of our regulations to enforce than when you are dealing with a direct export of a box.
Page 27 PREV PAGE TOP OF DOC Segment 2 Of 2
Now, I think when you get above—and we can argue this—I think when you get above 7,000 MTOPS, right now the volume of transactions and the capacity to string things together has dropped off to the point where I think we are at a controllable level. Will that last forever? No. Will it last more than a year or two? I don't think so. And that is one of the questions we have asked the consultants to study. That study is not complete yet, so I don't have an evaluation for you.
Mr. DELLUMS. Based on that answer, then—and I recognize that we have legitimate differences of opinion on these matters, that we have struggled over these things, so I can handle the disagreement—what do you think about my earlier statement that I recognize that we live in a global economy and it is a small grossly interdependent and interrelated world that we have a responsibility to assert some leadership in export control and sit down with other nations?
Someone much wiser than this gentleman once said that our nation-states are rapidly becoming yesterday's tribalism because there is some kind of way where we have to move beyond the boundaries of our own country to figure out how to gain control of these issues.
Secretary REINSCH. Dr. Wallerstein would like to comment on that.
Let me say I agree with you, Mr. Chairman, and we have been trying very hard to exercise that leadership. I referred to this issue in the context of the Wassenaar agreement, which is one of those multilateral arrangements. There have been proposals to Wassenaar to decontrol computers at much higher levels than what we agreed to, and we did not agree to those proposals. We are trying to exercise the kind of leadership that you want.
Page 28 PREV PAGE TOP OF DOC Segment 2 Of 2
I think Mitch wants to add to that.
Mr. WALLERSTEIN. Yes, I would like to comment, Congressman. I think you are absolutely right. I think that if we do not exercise leadership, no one will on these issues. We are the technology leader, particularly in this field.
But we also have to recognize that with the end of the cold war, the unifying force that caused the technologically advanced nations to band together and to come to, reasonably readily, agreement on the need for controls and on control levels has begun to deteriorate and to deteriorate fairly substantially, so that if we do not present a credible case for controlling, whether it is computers or any other advanced technology, at a level that these other governments also find credible, they are not going to be willing to sign up to it.
So we have constantly sought to balance our requirement to lead, and we clearly recognize that, with the needs to present a credible and technologically defensible case.
Mr. DELLUMS. With your permission, I would like to go further. We made an effort in this legislation to say, look, if the issue is threshold, then let us give the President the capacity to raise the threshold to whatever is deemed appropriate. And whatever that threshold is, and I am not an expert in this area, we simply said whatever the threshold is, you ought to be able to control it.
So I would like you to comment, one, what about our efforts to give the President the flexibility to raise the threshold, and once the threshold is placed at a higher level, would you not agree with me that we need to assert some kind of controls? And if you disagree with this, what kind of controls would you put in its place?
Page 29 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. Let me say first, Mr. Chairman, that we appreciate the efforts of you and other members of the committee to build that flexibility into the provision, and I think it was a very important element. In our desire to discuss the negative parts of the amendment we sometimes leave out the positive parts, and that was one of them.
Nevertheless, I think I could say that two problems remain. One is the waiting period that has to ensue after a change is made.
Mr. DELLUMS. From that point, may I ask a question, and maybe a technical question?
My thought was, as we were discussing this matter, in my own mind, that we did not necessarily need to use up the 180 days; that if there were extenuating circumstances, the need to move forward much quicker, then we did not have to sit there and wait for 180 days, if it needed to be done in a shorter period of time.
Maybe you or staff, as you looked at it, does the language, as you perceive it, is it so rigid that we do have to use the 180 days or can we slow it down and shorten the timeframe if necessary? How do you read that?
Secretary REINSCH. From the standpoint of the people who would have to administer it, Mr. Dellums, I would tend to view it a little bit like a situation involving an Appropriations Committee reprogramming request, in the sense that we would wait the full time unless the committees of jurisdiction chose to tell us in advance of that time that they had no objection. And if they had no objection, then we would go forward at that point. We would defer to the committees——
Page 30 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. DELLUMS. Could you—I am sorry, I did not mean to step on your comment.
Could you conceive of a situation where you would come to the committee and say we cannot wait 180 days, there are some extenuating circumstances here that we believe warrant moving quicker? Do you think as you read the legislation that that flexibility is there?
Secretary REINSCH. Well, I think we could always come to the committee. And I think as a practical matter, if the committee were content with what we wanted to do, within the context of the waiting period, I think we would feel comfortable going ahead and doing it.
But if I can add one thing. What really worries me about the amendment, and has come to worry me more in the last couple of weeks since I have watched the study underway evolve, is the amendment may be locking us into a method of controlling computers that we may all collectively decide at a later point is not the most appropriate one.
We may be heading toward a world in which we are caught essentially between two dilemmas, and that is, on the one hand if you want to eliminate all of the national security risks that are identifiable with computers, you need to control them at such a low level, down to the common PC level, that it really is absurd, and you would be talking about millions of licenses and a lot of money and you would still be ineffective.
Alternatively, because of the technological developments that Dr. Wallerstein and I have both described, we may, at the upper boundary, be reaching a point if we try to control solely on the basis of MTOPS across the board, related to what is available overseas, because of scaleability and other things, we may have to set a level that is so high we will be letting some things go we should not let go.
Page 31 PREV PAGE TOP OF DOC Segment 2 Of 2
What that suggests, and we do not have either conclusions here yet, nor do we have evidence, what that suggests is something that Mitch and I will want to work on, which is giving some thought to whether there is a more finely grained way we should control these things, and that it should not be simply a question of MTOPS level, where there are other capabilities of specific computers or even specific products that ought to deserve deferential treatment for national security reasons. And one of my concerns is the amendment would make it much more difficult for us to do that.
It really pushes us into the same construct we are in now, four tiers of countries and a unitary MTOPS, end MTOPS levels. We have not come to the conclusion that that is unwise at this point. What I see developing is some evidence to suggest that it might be.
Mr. DELLUMS. I think that is important testimony, and I think that is something we need to look at and continue to dialog around.
Just one last thing. I do not want to dominate all the time, Mr. Chairman, because I could ask a number of questions, but just one final thing.
Of the five tier III countries, three of them, India, Pakistan, and Israel, is it not true that they have their destiny in their own hands in the sense that they can remove themselves from tier III by signing up to the Nonproliferation Treaty?
Secretary REINSCH. I am not entirely sure about that, Mr. Chairman—Mr. Dellums, I am sorry.
Page 32 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. DELLUMS. You said that a few times. Former chairman. But in this business they always talk about your last highest rank, so it's cool, if it's all right with——
Secretary REINSCH. No disrespect intended.
Mr. DELLUMS. See, there are two of us here, Chairman Dellums and the ranking member.
Secretary REINSCH. No disrespect intended to the current chairman, Mr. Spence.
I really have to get further guidance from the committee about that. As I understand the bill, the language created two conditions for removal, one related to a category of countries that are listed in the CTBT negotiations in article 14, annex 2 to that treaty. And the countries you have questioned are listed there. Precisely what they have to do to get off that list, I am not sure.
Obviously signing the Nonproliferation Treaty would be one item, but if there are different criteria, I would defer to your judgment about that.
Mr. DELLUMS. OK. Mr. Chairman, we need to find that out. As you know, in the discussion on the floor a few days ago, the chairman of the Committee on International Relations disagreed on this issue. So we need to find out definitively. My thought, as we had moved along in these negotiations, was that these three countries had destiny in their own hands; that they could get off tier III on their own, but we need to clarify that.
Page 33 PREV PAGE TOP OF DOC Segment 2 Of 2
You have been generous, Mr. Chairman, and I would yield my time back.
The CHAIRMAN. Thank you. Mr. Bartlett.
Mr. BARTLETT. Thank you very much. Apparently, we are caught on the two horns of a dilemma. On the one hand, I think that it can be reasoned that everybody benefits, both economically and militarily, if we are permitted to sell, if our manufacturers are permitted to sell whatever technology can be purchased elsewhere, elsewhere in the world. If it can be purchased elsewhere then I have the feeling we ought to be able to compete in that marketplace.
You can make the argument not only does it benefit us economically but it also benefits us militarily because a broader, more vigorous industry will provide ever better and better computing capabilities. So that is the one horn of the dilemma.
The other is that in today's world the computer utility is the combination of hardware and software. And with relatively small computers, laptops for instance, with the technology of networking and paralleling that is made available by ever better and better software, you can take these smaller computers and can compound them until you now have what would be reasonably defined as supercomputers.
Clearly, we do not want to give any of our potential enemies a technology that they could not acquire elsewhere. This, as you have pointed out in your testimony, is a very difficult dilemma that we are in. I am not sure how we meet our responsibility as a committee while permitting you the kind of flexibility that you need in regulating these exports.
Page 34 PREV PAGE TOP OF DOC Segment 2 Of 2
We have sought to do that in this legislation. We were very much concerned that the bill before our input would not have adequate control over export of computers. I guess my question is, since the computing utility is really a combination of the hardware and software, how do we regulate that, since I do not think these export controls relate to software?
Secretary REINSCH. We do control certain software for export, Mr. Bartlett, and subject to other provisions of our regulations, and we could do that.
As I am sure you are aware, enforcement in that area because of the nature of software is much more difficult. We are talking about a product that can be downloaded on the Internet or shipped out of the country via telephone, not to mention one's pocket, going through the airport metal detector without setting it off. Shipping a high performance great computer is different. It is a big box. And when it leaves the country, people know about it. So enforcement is difficult.
And I could not agree more with the dilemma, the explosion of software capabilities, particularly in the interconnect area, which is a new thing, complicates the task, because what you have now is software in existence that will provide functions that additional hardware was formerly able to provide, and that makes things much more difficult.
There is, in addition, another dilemma I want to offer up, just to complicate the committee's life further, and that is that the Pentagon, or the Department of Defense, is an important consumer of computers. They need high performance computers for their own purposes, as Dr. Wallerstein has pointed out. They are not, however, an important enough consumer of computers to by themselves keep these companies in business. These companies need civilian sales for their survival.
Page 35 PREV PAGE TOP OF DOC Segment 2 Of 2
One of the things I would urge the committee to do is do a study of the health of this industry. There has been a lot of consolidation in this industry in recent years. Not all of these companies are making as much money as they would like to make. One of the things that the Pentagon has to keep in mind, and one of the things that we worry about as well, is the health of the domestic industry, which increasingly is a function of exports, because that has a direct bearing on what the Pentagon is able to buy.
So that is an additional piece of the dilemma. If we control too tightly, we may be putting people out of business, and that will come back to bite the Defense Department, too.
Mr. BARTLETT. I mentioned that in my initial remarks, that we benefit militarily when we permit more export sales because we now have a more vigorous industry which will better produce the type of equipment that we need in this country.
Secretary REINSCH. Exactly.
Mr. BARTLETT. It appears to me as this technology continues to grow, that in the future we will approach a point where it is an exercise in futility to try to control exports, because there will be so much available and so much capability to synthesize from smaller units enormous computing capabilities through networking and paralleling.
Secretary REINSCH. That is what worries us. We are not at that point now, however.
Page 36 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. BARTLETT. Thank you very much.
Secretary REINSCH. Thank you.
The CHAIRMAN. Mr. Pickett.
Mr. PICKETT. Thank you, Mr. Chairman. You have mentioned in your comments this morning that putting a standard into law is rather inflexible, or makes a rather inflexible program because the technology is changing so rapidly.
I think some reference was made to the fact that it would be better to establish perhaps some kind of way to adjust this other than through legislation because it takes so long to change things that get into law.
Secretary REINSCH. Yes, sir.
Mr. PICKETT. A point that I do not quite understand, you mentioned about actually physically exporting equipment, but what about communications and having access to high performance computers that do not leave the country?
We have heard lots of stories about hackers getting into computers at the Defense Department and other government agencies. Once you get in, I take it, if somebody is smart enough and knows how to operate the equipment, they do not physically need to have the computer, it is out of the country, to manipulate it and use it. Is that a fact or not?
Page 37 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. That is a very good point, Mr. Pickett, and I am glad you brought it up.
One of the points I have made in other contexts has been, frankly, if I were the Chinese, and I wanted to acquire a high performance computer that I knew the American Government did not want to sell, I would not make the effort. I would set up a sub onshore, out on the West Coast someplace, buy two or three of these things, it is not on export, do the computation here, ship the data back by modem over the phone or take it back in a suitcase. Nobody would ever know, and there is no export involved.
This is one of the not unique but peculiar features of this technology and this market. What the computers do, they are not like machine tools, if you will, and they are not like chemical precursors or biotoxins. They do not make other things directly and they do not—they are not weapons directly. They make information. They make data. And that is the most highly transportable commodity in existence.
Now, with respect to domestic hacking and that kind of problem, I think that is worthy of further study by your committee. There is an organization called the President's Commission on Critical Infrastructure Protection, which has issued its report on precisely this problem, cyber attacks on domestic infrastructure.
I know there have been two hearings held in the Congress so far, one in the House Science Committee and one in the Senate Judiciary Committee. This is very much an issue that your committee may want to take a look at as well. The commission makes a large number of recommendations, which they will be submitting to the President fairly shortly.
Page 38 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. PICKETT. Well, is it fair to say, even if we design a very strict, or an export control regime that is ironclad, there could still be considerable leakage through the use of communications and not taking hardware out of the country potentially?
Secretary REINSCH. By a committed adversary, yes, that is correct.
Mr. PICKETT. Thank you. Thank you, Mr. Chairman.
The CHAIRMAN. Thank you. Mr. Gibbons.
Mr. GIBBONS. Thank you, Mr. Chairman.
Mr. Reinsch, the computer that Silicon Graphics sent to China, where is that computer right now?
Secretary REINSCH. Let me preface my comment, Mr. Gibbons, by saying this is a matter that is under investigation, being led by the Justice Department, so there is a limit on how much I can say about it.
I believe the computer, that the particular SGI computer is located at its intended destination, the Chinese Academy of Sciences.
Mr. GIBBONS. How long ago did you request it be returned?
Page 39 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. I do not believe that we have requested that that one be returned. You may be thinking about the Sun Microsystems computer that was exported to a military end user in China. We did ask for the return of that computer, and that computer was, I am informed by the company, that computer was returned to the United States on November 9.
Mr. GIBBONS. Was the Silicon Graphics computer that went to the Chinese Academy of Sciences under the export control licensing requirements of the Department of Commerce?
Secretary REINSCH. That is a judgment that we have to make in concert with the Justice Department, which is currently investigating the circumstance of that export.
Mr. GIBBONS. What was the speed of that computer?
Secretary REINSCH. Pardon me.
Mr. GIBBONS. What was the MTOPS speed of that computer?
Secretary REINSCH. I think I can give you that one. If you will bear with me a moment while I get my chart. I have so many charts that—I believe that one was in the range of 5,000, and I will have to get you the exact number.
[The information referred to can be found in the appendix beginning on page 73.]
Page 40 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. GIBBONS. 5,000 MTOPS.
Secretary REINSCH. I believe so, yes. It was under seven. Everything we have been talking about in this entire episode, if you will, has been between 2,000 and 7,000.
Mr. GIBBONS. What is the floor threshold of the licensing requirement?
Secretary REINSCH. With respect to China?
Mr. GIBBONS. Yes.
Secretary REINSCH. With respect to China, we would require an individual license, which means advance approval, above 2,000 for military proliferation related end users and above 7,000 for civilian end users.
Mr. GIBBONS. Did Silicon Graphics apply for an export license of this computer?
Secretary REINSCH. No.
Mr. GIBBONS. Do you feel, just today, telling us what you have told us, that Silicon Graphics violated the law in this regard?
Page 41 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. Well, I have been asked that question before, and I think the best way to answer that is I really don't want to comment in advance of the Justice Department's investigation because I don't want to complicate what they are trying to do.
Mr. GIBBONS. What about IBM's sale of HPC's to the Chelyabinsk-70 facility?
Secretary REINSCH. There are, again these are matters under investigation by the Justice Department, so there is a limit to what I can comment on. The IBM sales were not to Chelyabinsk but to a different Russian facility known as Arzamas 16. And according to published reports, there were—you could essentially group this into two categories: A high performance computer that was shipped in one transaction and 16 other computers that were shipped in another transaction, which may or might not have been linked together.
Our understanding with respect to the 16 is that their individual MTOPS rating, the MTOPS or the CTP rating of each one was, I believe, 275.9 MTOPS each. Our engineers have indicated that because of the different nature of interconnect devices, they could not tell whether the computers had been linked together in a way that allowed them to operate collectively without examining them.
So I could not tell you if there is a collective force there or not.
Mr. GIBBONS. But it is possible to connect each one of those together to create a supercomputer, is it not?
Page 42 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. Yes, it is possible. Whether it has happened is what I cannot tell you. I would also say, though, that what you get is something less than the sum of the parts when you do that. That is that the total MTOPS equivalent would be something less than 16 times 275.
Mr. GIBBONS. Sure, but it would still fall within the minimum floor level.
Secretary REINSCH. It would fall between 2,000 and 7,000; that is correct.
Mr. GIBBONS. Precisely. Understanding all that, what changes have you instituted at the Department of Commerce to ensure that these high performance computers do not reach the hands of nonauthorized tier III countries following the revelations that we have before us today about Silicon Graphics and IBM's shipments? What changes specifically have you instituted?
Secretary REINSCH. Well, what we have done I alluded to in the statement that Ms. DeBusk read for me. What we have already imposed upon the companies, by virtue of the President's decision to change the policy in 1995, was a comprehensive recordkeeping and data collection requirement. So they are required to keep records of all computers above 2,000 that they export to all destinations.
We are obtaining and apply that information at quarterly intervals, and we are going through each of the computers that they notify us of and making judgments as to the nature of end use, whether further investigation or action is warranted.
Page 43 PREV PAGE TOP OF DOC Segment 2 Of 2
In the Sun Microsystems case in China, for example, that was a product of our investigatory work. We concluded that the computer should not have gone there. We asked for it back; we got it back. How it got there is still a matter of investigation, and there may be penalties, but that is a Justice Department question. We will take those steps in other cases as they occur.
I indicated that with respect to tier III countries we have, on the basis of the information we have received, requested 22 postshipment verifications of other computers. Seventeen of those have been completed so far, and they have been favorable, which means we did not find anything that would warrant further investigation.
Mr. GIBBONS. It just seems to me that this policy allows for the cow to get out of the barn, and then you decide to close the door, because your policy changes of 1995 did not stop Silicon Graphics' shipment, did not stop the IBM shipments. It is a review process after the fact.
We now have these terrorist countries out there, or countries that should not have these computers, to have them in their hands. So I am concerned that the policy is one of closing the door after the cow has got out of the barn.
Mr. Chairman, I thank you very much for your time and yield back.
The CHAIRMAN. Ms. Harman.
Page 44 PREV PAGE TOP OF DOC Segment 2 Of 2
Ms. HARMAN. Thank you, Mr. Chairman. I would like to thank the witnesses for excellent testimony. I think it was carefully drawn and very sensitive to the balance here that we are trying to strike.
I was one who did not agree with what our committee did. I voted ''no'' in committee, and I filed separate remarks because I think that while we tried hard and while our compromise is a great improvement over the original form of the amendment offered to the Defense authorization bill, we did not get it right. And as I listen to this testimony, the first question that occurs to me is do we have a regulation problem or do we really have an enforcement problem?
Secretary REINSCH. I have always believed that it is primarily an enforcement problem, Ms. Harman. And I appreciate the question.
I think the policy has been appropriate. The policy has been calibrated to, in light of both availability and our national security needs. We do have companies that appear to have violated the law, as happens with other laws, and we are determined to catch them. And I believe to follow up Mr. Gibbons' comment, probably the best deterrent effect is effective prosecution.
Ms. HARMAN. Does the Defense Department have a comment on this?
Mr. WALLERSTEIN. No, I believe that Mr. Reinsch has adequately characterized the overall U.S. Government position on this.
Ms. HARMAN. My second question concerns another subject, but it raises similar issues, and that is the regulation of encryption exports. I see you break out in hives, Mr. Reinsch.
Page 45 PREV PAGE TOP OF DOC Segment 2 Of 2
But my question to you is, are there similarities between these two problems and should we be thinking about them in tandem rather than separately?
Secretary REINSCH. Yes, there are; yes, you should; and the last time I was here, I think the irony of me telling this committee that encryption controls were a good thing and supercomputer controls were not a good thing did not escape the members of the committee. At that time I did not point out the irony of the committee saying that encryption controls were a bad thing and computer controls were a good thing was also ironic.
There are some similarities in this issue. The biggest one, it seems to me, is also the question of enforceability. Encryption is something that is transmitted through the—either downloaded on the net through software, which is one of the most invisible things you can export. It is very difficult.
There are, like with computers, clear national security implications to the actions that we take, and I do not want anyone to believe, as I think was clear in my exchange with Mr. Dellums, that the Department of Commerce or the administration opposes controls on computers. We do not. Neither do we oppose controls on encryption, as you well know.
Ms. HARMAN. Well, I just would have a final comment, Mr. Chairman. As you know, I sit on this committee and on the Intelligence Committee. I also represent a district in California which is finally diversifying, weaning itself from Defense dependence and going into other high technology activity which depends for its future on exports. So I am trying to figure this out.
Page 46 PREV PAGE TOP OF DOC Segment 2 Of 2
I am glad that we have witnesses of this caliber to help us figure it out, and I would just like to suggest that we work closely together during the next session of this Congress to try to get it right.
Secretary REINSCH. We are happy to—if I may say, Mr. Chairman, historically, BXA and the Department of Commerce's relationship has been primarily with its authorizing committee, the International Relations Committee. We are happy to have a close relationship with your committee. We are happy to consult. I believe we have done so. I believe we have a good working relationship with your staff, although they can comment as they wish, and we are very much interested in continuing that relationship and are happy to come up and see you under any circumstance at any time. And I am sure since you are his authorizing committee, I am sure he will tell you the same thing.
Mr. WALLERSTEIN. I tell you the same thing.
Ms. HARMAN. Well, I conclude my comments, Mr. Chairman, but I would certainly like to take Mr. Reinsch up on that offer, because I think we need to get a lot wiser on this subject.
Even assuming that the Defense authorization bill is signed and this new provision becomes law, I still think we need to get wiser, and perhaps some corrections down the road would be in order.
Thank you, Mr. Chairman.
Page 47 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. Thank you, ma'am.
Mr. Hunter.
Mr. HUNTER. Thank you, Mr. Chairman.
Good morning, folks. I have looked at the provisions that were passed by the conference report, which were largely the product of the chairman and ranking member's work here on the committee, and I want to go over that report and just ask you, as we go over each section, what problem you have with it.
Mr. Wallerstein, the import I have gotten from you and from all of our witnesses is the idea that good judgment is needed here; that if smart people in the administration will simply look at the speed of modernization with respect to computing capability and the foreign availability that is present now and projected to exist in the future, that a lot of these computers that are 2,000 MTOPS or around that area will be waved through by anybody who is a rational, reasonable person.
Is that your testimony?
Mr. WALLERSTEIN. I think what I was trying to convey was that we need to take a comprehensive approach which looks at the availability, that looks at the controllability of the particular machine.
Page 48 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. OK. And so what we have here is a provision—No. 1, section 1211, says that any proposed export will receive a 10-day review. That is export of computers more than 2,000 MTOPS. A 10-day review by the Secretaries of Defense, Commerce, Energy, State, and the Director of the Arms Control and Disarmament Agency. In this case, all those representatives of those agencies of the Clinton administration. That means Bill Cohen, Secretary of Defense, will be looking at this for 10 days.
Now, if what you have told us is true, that any rational, reasonable human being will wave this thing through on the basis of all the things you have just said—foreign availability, modernization that is taking place around the world, and the need for our industry to maintain its balanced level—why are you afraid of your boss in DOD having 10 days to look at it? Why is that not reasonable?
Mr. WALLERSTEIN. We certainly wish to look at every single export and, in fact, under the terms of Executive Order 12981, which was promulgated in December 1995, we obtained the right to do that.
I think the concern that has been expressed—and Mr. Reinsch has articulated this earlier—is that the 10-day period does amount to, in effect, a double-licensing requirement. Because it says that they have to be submitted and reviewed and then, if a license is required, come back in and go through the actual licensing period.
The bottom line for defense——
Mr. HUNTER. No, no, wait a second. Let me stop you on that. The only time licensing is required is if President Clinton's Secretary of Defense, Mr. Cohen, or the Secretary of Commerce or the Director of Arms Control and Disarmament raises a red flag and says, I have a problem with this.
Page 49 PREV PAGE TOP OF DOC Segment 2 Of 2
Now, you have just told us, the import of all your testimony has been that reasonable people, exercising good judgment, are going to wave most of this stuff through. So why are you afraid of your boss and the boss of Commerce and Arms Control looking at the thing for 10 days?
Mr. WALLERSTEIN. I am certainly not opposed to that at all. I would only be opposed to taking up his time twice on the same subject. So I would like to have a system that is efficient.
Mr. HUNTER. OK. Are you saying this is going to—that your boss, Mr. Cohen, does not want to have the authority to look at this thing for 10 days?
Mr. WALLERSTEIN. The position that the administration has taken—and my boss is part of that administration—is that we believe we have sufficient authority under the existing Executive order.
Mr. HUNTER. So you would be opposed to giving Bill Cohen the right to look at this for 10 days and raise an objection if he so wishes?
Mr. WALLERSTEIN. No, I did not say that.
Mr. HUNTER. Well, that is what the section says. So we have to take the section——
Page 50 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. May I make a comment, Mr. Hunter?
Mr. HUNTER. Sure.
Secretary REINSCH. The administrative burden, if you will, that would be particularly troublesome in that provision would be the referral of these licenses to the Nonproliferation Center for an end user comment.
I would expect that most, if not all, of the Cabinet Secretaries named would want to get the NPC's judgment on the prospective end user. The volume and the time limits of the provision are such that it would be very difficult for the NPC to meet those constraints.
Mr. HUNTER. So what you are saying is, it is very difficult to find out who the end user is in many of these cases?
Secretary REINSCH. No, I am saying that it is very difficult for the NPC—it is going to be very difficult, with the resources the NPC has, to deal with the volume of these requests in 10-day periods.
Mr. HUNTER. Well, so, if it is a routine 2,000 MTOPS sale, and everybody has got it, why should this put an undue burden on people of good judgment?
Secretary REINSCH. Well, if everybody has good judgment, it might not.
Page 51 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. Well, we are talking about the Secretaries of the Clinton administration.
Secretary REINSCH. They are all people of great wisdom and good judgment, Mr. Hunter, who would act on staff advice.
Mr. HUNTER. In a way, you are making our point here, Mr. Reinsch. You are saying common sense tells us a lot of this stuff should be waved through and a bureaucratic hurdle should not be in the way.
We have not placed a bureaucratic hurdle in the way; we have given a 10-day review, which is a very short period of time, for the combined leadership of your administration. And in their good judgment, if this is stuff that can be bought literally off a Kmart shelf, this is not going to take Bill Cohen's staff 2 seconds to wave it through.
If, on the other hand, it is dangerous or it poses a danger to the United States, he is going to raise a red flag and then, rightly so, if, in the opinion of Bill Cohen, selling a particular supercomputer to a particular party is a bad thing, would you not agree that that sale should be explored? Would you agree with that?
Will the record reflect, Mr. Wallerstein is nodding his head yes.
Then what is wrong with having a provision in the bill that gives him that power to be cognizant of all the sales so that he can pick out and red flag the sales that look bad? What is wrong with that?
Page 52 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. WALLERSTEIN. There is nothing wrong with the idea in principle, Mr. Hunter. As I said, we are opposing it more as an administration efficiency argument.
Mr. HUNTER. Let me go further. I have talked to some of the executives of these companies that have made sales. They are not confident, they are not sure of what they are doing. They are confused and a little bit bitter for the administration sending them what I think they feel are mixed signals.
For example, when they get an order from a theoretically benign user, the agriculture department of a particular country, knowing it is illegal for them to sell if the invoice is from a military department of that country, and yet having the common sense to know that probably the military department of that particular country can access any agency it wants, including the agriculture agency and the weather agency's resources, they know darn well what they are sending to that country, that piece of paper notwithstanding, will be accessed by the military. And yet we are asking them on this honor system that is employed now to make a determination that that is a benign end use.
And I have talked to some of them. They know it is a game.
Why is it in our interest to present this fiction to the private sector and expect them to exercise some sort of judgment that is better than the judgment of, for example, our collective Secretaries? Don't you think that is a disservice to them?
Page 53 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. When the element of the President's policy that made a distinction between military and civilian users was made, my recollection is that the reason for that distinction was not because we felt we could effectively control within the entire country, because of the reasons you have said, but for actually the reasons that Mr. Dellums raised in his questions, which is to demonstrate leadership and to demonstrate it was not U.S. Government policy that those entities should end up with those items and have that capacity.
But we did not have any illusion at the time that we could effectively forestall committed adversaries from obtaining the capacity.
Mr. HUNTER. Let me ask you another question. When we had the sale to Communist China—and, incidentally, we had some of our smartest guys from our weapons laboratories give us what we thought was a pretty dispassionate analysis, at least I perceived that, on what the import of the sales to Russia and to China were—the essence of their recommendation, to me at least, was that the sales to China, to Russia, the supercomputer that Mr. Mikhaylov boasted about that he received and apparently wasn't so easy to get off the Kmart shelves, because I think he had a press conference when he got the one from the United States, but that sale was something that was very easy for him.
Or that sale to Russia was something that Hazel O'Leary said originally she would not make. Mr. Mikhaylov sent back a letter saying, in fact, that Mikhaylov asked for it. Hazel O'Leary said, you are not going to use it for military use, et cetera, and they said, yes, we are going to use it for stockpile stewardship, and yet we ended up making that sale.
That sale was acquired, I think, from Silicon Graphics. As I understand, the person that made that sale for Silicon Graphics was punished for doing that.
Page 54 PREV PAGE TOP OF DOC Segment 2 Of 2
Now, you have made the statement that the present policy apparently is a good policy. The impression that I got from Silicon Graphics and from the company that made the sale was that there was a lot of confusion and there was a lot of resentment because there was what they thought was an ambiguous policy that ended up with an employee being punished for making the sale.
Do you really feel comfortable with the policy that existed when that occurred?
Secretary REINSCH. I think I would say two things about that.
First of all, when something goes wrong, Mr. Hunter, there is always a lot of finger-pointing. And all the parties in that particular case have said different things at different times, most of them in an effort to avoid the blame, so I am not surprised at the statements that you referred to.
Mr. HUNTER. I am talking about what really happened. And I just wondered, Mr. Reinsch, if that employee was punished for making that sale under, at least to them, what was not a clear policy, wouldn't it have been a lot better to have had the chairman and the ranking member's policy that said, if you are going to make a sale like that, you have to let the Secretary of Defense and others know about it, and they either sign off or they don't sign off?
Then government, our Government, with our security responsibilities, has made the determination. The company has not been forced under this honor system to make a decision which may or may not reflect adversely on them and on our country.
Page 55 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. WALLERSTEIN. If I may intervene here, I would remind you, Mr. Hunter, I think you are aware of the fact there had been previous license applications submitted for exports to those same destinations, and the U.S. Government had turned down those license requests, so that we had already reached a view on the desirability or lack of desirability of exporting to those destinations.
The claim that the exporter did not know—and again I don't want to get into too much of the details, because this is still under investigation, but the claim that the exporter did not know what these end users were is like claiming that you don't know what goes on at Los Alamos or Livermore. It is well established what the purposes of these laboratories are.
So given that we had already established a precedent and that the purposes of those facilities are well known, I don't put much stock in the claim that they just didn't know and they were confused.
The CHAIRMAN. We are going to break for a vote right now and we are coming right back.
[Recess.]
The CHAIRMAN. The meeting will please be in order. I apologize for the length of that vote, but they had two votes, so we had to stay for that, and we will try to hurry on.
Page 56 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. Skelton is recognized.
Mr. SKELTON. Mr. Dellums asked all the questions that I had.
The CHAIRMAN. Well, anyway, I was just thinking about pointing out, I was going to leave it for somebody else to do, but the problem you have in sending these computers to other countries is the end-use aspect of it. It is difficult to say what the end use will be, especially in relation to China, because they will not even let you know what the end use of it is, and it ends up being in the military. That is the problem we have.
And these middlemen, too, can sell to other people. What about Iran and Libya and people like that being able to get hold of them through middlemen, even if we do not try to shut it off to the direct source? How do we handle that in approving these export licenses?
Secretary REINSCH. Well, you are quite correct, Mr. Chairman, in pointing out that this is a problem. Our regulations are very clear about this. We do not permit these things to be exported to those destinations. And whether it is a direct export or a re-export, that is, from some middleman, the regulations are clear that they would have to seek a license, which would be denied.
Now, it is a fact of life, as I have told the committee in the past, that when an item—not just a computer, any item—leaves our shores, we lose some measure of control over it. When it goes to a destination that is an ally destination, a member of Wassenaar, a member of one of the other regimes, we have a high degree of confidence that those countries have parallel export controls to which these items will be subjected.
Page 57 PREV PAGE TOP OF DOC Segment 2 Of 2
When it goes to a destination where we don't have that level of confidence, because they do not belong to the regimes, it is a bigger problem. Our law is clear, it is just harder to enforce and we rely on the intelligence community to do the best we can as far as gathering information.
We send—Ms. DeBusk sends our agents out periodically on lengthy safeguards trips where we will do postshipment checks on computers in place, but also meet with officials in countries that are not members of multilateral regimes and where we can talk to them about improving their own controls and their own supervision of what passes through their borders.
The CHAIRMAN. Well, do you get this information that you send to the people from our intelligence agencies and DOD, and their input? That is kind of what we are talking about doing here in this 10-day period.
Secretary REINSCH. We would like to think we work closely with all of these various parties. As far as intelligence is concerned, we work primarily through the Nonproliferation Center which, as I understand it, speaks for the entire intelligence community and receives input from the entire intelligence community. I think we have a very good working relationship with them.
I know when we came up to brief your staff on the Russia-Chinese situations, we brought their representatives with us to comment on the specifics; and I am sure you or your staff can contact them directly either for additional information or to comment on our working relationship. I think it is a good one. They perform a very important service for us.
Page 58 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. Mr. Hunter, you were in the middle of some questions a while ago. Thank you.
Mr. HUNTER. Yes, thank you, Mr. Chairman.
Has China made diversions from their so-called benign agencies that have received supercomputers to military applications? Have there been diversions there?
Secretary REINSCH. I think, Mr. Hunter, the Sun case would probably be considered a diversion in the sense that it did not end up at what we were initially told was the destination. It ended up at a different destination. Who was responsible for that diversion, whether it was the Chinese or another party, is the subject of investigation.
Mr. HUNTER. Where did that go?
Secretary REINSCH. It went to the Changsha research facility in Changsha.
Mr. HUNTER. What was that? How many MTOPS was that?
Secretary REINSCH. Two thousand seven hundred eight, I
believe.
Page 59 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. That is a facility that is involved in weapons development.
Secretary REINSCH. I don't want to be more specific than to say it is a facility that is involved in military research and development.
Mr. HUNTER. Was that a sale made under a license?
Secretary REINSCH. No.
Mr. HUNTER. It was sold to an ostensibly benign user, and it ended up being shifted?
Secretary REINSCH. That is correct.
Mr. HUNTER. Was it ever—what was the connection—what was the registration, if you will, or the recordation of that sale, if any, with the U.S. Government? What records do we have on that?
Secretary REINSCH. When we received the data—and again this Sun case is still under investigation, so I don't want to go into too much detail—when we initially received all the data with respect to all the companies, this particular one was indicated as having gone to a different destination, the name of which I do not remember offhand, but it is not particularly important.
Page 60 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. But a benign——
Secretary REINSCH. Well, not necessarily. It was an entity in question that we did not recognize, so we consulted with the intelligence community, as we do on these. They did not recognize it either; and based on that, we went back to the company and made some further inquires, at which time the company told us that that was not the final destination, that in fact there was a different final destination, which was Changsha. And that is a facility that we have some familiarity with.
Mr. HUNTER. OK. When we first had the blowup over supercomputer transfers, and we asked you folks to come in and brief us, as I recall, there was some state of confusion as to exactly what the number of supercomputer transfers was. It was some weeks before we got the number, 47 supercomputer transfers. Is that accurate? You folks did not have that at your fingertips when we first asked for that.
Secretary REINSCH. When you first asked, no. That is correct. You first asked, I think, in February.
Mr. HUNTER. So when did your folks become cognizant of the fact that there were that many United States supercomputers in China?
Secretary REINSCH. When we received the data.
Mr. HUNTER. Well, who did you receive it from?
Page 61 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. Well, the President's policy required the companies, from the beginning, to keep the data. We did not ask them initially to submit it to us until roughly February, 1997.
Mr. HUNTER. So the point is, at that time, until you asked, until you did a little investigation on it yourself, you had no idea there were 47 supercomputers in China?
Secretary REINSCH. That is correct.
Mr. HUNTER. So it goes without saying you had no idea whether there could have been supercomputers being used in military programs or not? You simply did not know?
Secretary REINSCH. Of course, you can make that—we did not know there were 253, or whatever it is, in Germany. We did not have that data.
Mr. HUNTER. In a way, Mr. Reinsch, I think you have made our point. We think it is kind of important to know, with all of the hundreds of thousands of people we have dedicated to accounting on very simple things, such as nuts and bolts in our Department of Defense, it does not seem like a huge allocation of resources to know where 47 supercomputers are, and it does not seem to me that there should then be—it puzzles me as to why there is such massive resistance in the administration. I mean, even for things like notifying your own boss, the Secretary of Defense, for 10 days, assuming he is a reasonable guy and he is going to wave it through, notifying him that the sale is about to be made; it is like we do not want to know.
Page 62 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. Well, I would not agree with that, Mr. Hunter. I think if volume were going to stay at 47, we would not be complaining so much. We project a big increase in volume worldwide.
Mr. HUNTER. Let me ask you another question. How do you view the nondefense agencies of the Chinese Government? Do you view them as independent, or do you view them as an agency that will transfer, on demand, technology to other agencies? Maybe Mr. Wallerstein could assist there.
Mr. WALLERSTEIN. It is clear, Mr. Hunter, that the organization of the Chinese Government is different than, for example, our own Government in terms of how some of these programs are managed. Some of what you are asking, of course, falls in an area of intelligence, and I would join Mr. Reinsch's suggestion that you may wish to consult further with the NPC on that matter.
Secretary REINSCH. They do have a view on that matter, and I think it would be worthwhile your having a discussion with them.
Mr. WALLERSTEIN. Of course, it is also true that as the Chinese have sought to diversify their economy and to develop more commercial activity, not state-run, not state-managed or owned, there is a growing nongovernmental or nondefense sector. So the challenge for us, from a regulatory standpoint, has been to find some way to have confidence that when we agree or permit an export, that it is going to a commercial end user not to a military end user, because I would remind you that there is an outright prohibition against assisting the military.
Page 63 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. Well, you made my point, Mr. Wallerstein. I have looked at these cases and I have asked your intelligence people to tell me if they could ever give me a case where the PLA in China was refused the access to technology from other agencies. They could not come up with any, indicating a total lack of what I would call independence or insularity on the part of the domestic agencies.
They all work for Beijing. Beijing has, obviously—it is obvious to all of us, every learned American would say that military capability is an important goal for Beijing. If there is a supercomputer that is going to give or assist that military goal, the idea that when the agriculture department gets the call or the weather department gets the call to send something over to the PLA, they are going to refuse to send it? I mean, we sit here with massive educations and sophistication and all the king's horses and all the king's men, but nobody with a lick of common sense is going to say that that is not going to happen.
We have created a fiction. We all know it is a fiction. We all know it is laughable. But we allow our domestic companies to engage, and I think it is a disservice to them to engage, in that fiction.
Do you know of any cases where a domestic agency has refused to access military technology or dual-use technology to a military agency in China?
Mr. WALLERSTEIN. Sir, I have no information one way or another on that subject.
Page 64 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. Then let me ask you another question.
Our intelligence people know which agencies tend to have a promiscuous record with respect to moving technology into the military, some more than others. I have had intelligence briefings, I am sure you have, on which ones are quite bad about that.
If I am the head of a supercomputer company in America that wants to sell, and I get an invoice that says, you are sending this to the weather department in China, and I do not have that background, why would you rather I make the decision as to whether or not that is a good sale than under our bill the ranking member and the chairman put together, giving the Secretary of Defense, who has access to that intelligence, to make that judgment? And you say judgment is an important thing here.
Why put that in the hands—Mr. Reinsch, for you, why put that in the hands of a private party? Because that is what we have done. And I have talked to some of these people, and they do not particularly like the responsibility that you have given them.
Secretary REINSCH. I think—let me approach this from two directions, if I may, Mr. Hunter. If your view is that the PLA will enable and will be inclined to access computers from any location in China, the logical conclusion of that view would be that we should not sell anybody in China any of these computers. That is not a use we subscribe to. We simply do not think it is that simple.
Mr. HUNTER. Except for one thing, and that is, at least if you have, if these computers have to go through a checkoff first with DOD and the other agencies, we at least are apprised up front. The Secretary of Defense and others can, through proper channels, notify China that the agriculture department has asked for this, it is going to be closely monitored; for example, we would like to have a letter from you stating that we can visit this computer on a regular basis, that we can check with our private people. You can put certain safeguards in that you cannot put in if you simply say the administration's position is they do not want to see it.
Page 65 PREV PAGE TOP OF DOC Segment 2 Of 2
I do not want my Secretary of Defense, Mr. Wallerstein—the import I get from him, he does not want to have the 10-day checkoff period, he does not want a review period.
Mr. WALLERSTEIN. I am sorry, Mr. Hunter, I have to interject. I did not say that and do not agree with that position.
Mr. HUNTER. OK. Well, we are all the world's greatest experts on our own opinion. I have been trying to get you to say you agree with our provision that gives you the 10-day checkoff, and I have not been able to get it from you. So that is my opinion.
But why put on private industry this decisionmaking power that because they do not have the classified briefings, they are not able to use total—they are not able to use a judgment that a Government official can use, having availed himself of all the information that is out there, including the classified information?
Secretary REINSCH. Well, you are making a good point. I think we will still disagree at the end of it, but you are making a good argument.
First, we do not put the entire burden on them. Our regulations are clear about this. If they have any doubts at all, they have an obligation to come to us and to pose that question to us, and some of them do that—not everyone, but some of them do that. We do a lot of outreach to try to get everybody else to do exactly the same thing if they have any doubts.
Page 66 PREV PAGE TOP OF DOC Segment 2 Of 2
Obviously, when something goes wrong—and I was trying to be tactful about this in my previous response, but obviously when something goes wrong, it is entirely normal for the company to blame the Government and the Government to blame the company. In fact, the company has an obligation if it has doubts about an end user to seek our advice and to get it.
In addition, as I indicated in my statement, we have begun—we decided to begin, sometime before this issue even came up, the process of listing unreliable end users publicly and notifying companies of their obligations.
Mr. HUNTER. You made one of my points. You said if the company has a doubt. You can have a jury out for months on what is a reasonable doubt and what is not. My point is, if they have not had the secret briefings and they do not know what is happening internally with the various agencies in a country like China or other places—and governmental officials do have those briefings, they do have the evidence from our intelligence agencies—how can they be expected to have the same judgment in making that decision as the agencies themselves, which can avail themselves of that classified information?
I mean, the CEO's of these companies do not get intelligence briefings on a classified level.
Secretary REINSCH. The CEO's of these companies are not the people making the decisions to sell the computers; the salesmen are, and sometimes their people on site, and sometimes they are people in country at subsidiaries. These are people that have access to a lot of information that we do not have access to.
Page 67 PREV PAGE TOP OF DOC Segment 2 Of 2
You are correct they do not have all the information that we have. They also have some information that we do not have. The best process is a collaborative one.
Mr. HUNTER. OK, the last question to you.
Why does it make sense—since you basically conceded that at least to some degree they are not as adequately briefed as people that have the intelligence available to them, which is our officials, why further vest this power, which can have enormous defense ramifications, on the party which is making the sale? That is, why give the veto to the sale to the party who stands to make a profit by making the sale? Does that not do that company a disservice, as well as the American people?
Secretary REINSCH. Well, if you look at the way export controls have evolved in this country generally, not even with respect to computers, the initial decision is always on the part of the exporter. He has to decide what his product is and he has to come to us and apply for a license for it or he has to come to us for advice.
Mr. HUNTER. But in some cases that is mandated by law. In this case, whether or not you have a reasonable doubt as to the benign characterization of the—or the benign quality of the end user is very subjective and very much subject to opinions and attitudes. That is not a legal requirement.
Secretary REINSCH. There are always judgment calls. There are other licenses, noncomputer-related licenses that we deny because the end user is unreliable and the companies' obligation to come in and consult with us is one that is not computer specific. It exists with respect to all items because of EPCI, the enhanced proliferation control initiative. We think this kind of collaborative process is the best way to get where we are going, to combine the information that the exporter has with the information that we have.
Page 68 PREV PAGE TOP OF DOC Segment 2 Of 2
If we are not going to trust the exporters at all, then you are envisioning a system of controls which is far more intrusive, far broader, and far more expensive than the one we have. One needs to draw the line somewhere.
Now, I suspect you and I would draw it in different places, but I do think at some point you have to say there is an element of business judgment that enters into this.
Mr. HUNTER. OK, let me ask you one last question. And, Mr. Chairman, thank you for indulging me here.
Would the Clinton administration go along with a 5-day notification period to the Secretary of Defense before a sale was made?
Secretary REINSCH. Well, I will make you——
Mr. HUNTER. And I am referring to—please refer to section 1211, because that is the section that requires prospective exports or re-exports of supercomputers with 2,000 MTOPS capability to countries of proliferation concern, or tier III countries, receiving 10-day review by the Secretaries of Defense, Commerce, Energy, State, and the Director of Arms Control and Disarmament Agency.
Would you go along with a 5-day review period? Is there any way you would allow the Secretary of Defense to see this before it goes out?
Page 69 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. I was going to offer you a different deal, Mr. Hunter, if you want to be practical about this. What I was going to suggest is, if you go back to 30 days on the waiting period, give us the flexibility the President wants in the tier III countries and drop the PSV provision, we can live with 10 days.
Mr. HUNTER. You want to have what, now?
Secretary REINSCH. Go back to 30 days on the waiting period for changing the MTOPS level, remove the limitations on changing the country lists, and drop the PSV, the postshipment verification provision, then you are talking business.
Mr. HUNTER. Why do you want to have—let me ask you a question about the MTOPS. Change in MTOPS levels. How many of those have been made?
Secretary REINSCH. Changes in MTOPS levels?
Mr. HUNTER. Yes.
Secretary REINSCH. Well, we have only had the policy—once.
Mr. HUNTER. So in the history of the world, how many times has the U.S. Government changed the MTOPS level?
Secretary REINSCH. It has changed twice in this administration, and we are committed to reviewing this policy every 18 months to 2 years. And the experts we have consulted with suggest that is the outside limit of which we should review it.
Page 70 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. So you only review the MTOPS—this is not something we do on a weekly basis, like looking at the markets and deciding we are going to adjust the interest rate. You only looked at this thing once, adjusted it two times, by definition, two times in a Presidential cycle.
So why is giving us 180, which is 6 months, to review that level, why is that mission impossible?
Secretary REINSCH. Well, first of all, as I said in my statement, I think 180 days is too long regardless of any construct. I don't see why you need 180 days, you only take 30 for changes in munitions controls.
Mr. HUNTER. My point is, you only changed MTOPS level how many times in the history of the world?
Secretary REINSCH. Once in 1993 and once in 1995. Yes, in previous administrations it has been similarly level.
Mr. HUNTER. How many times in previous administrations?
Mr. WALLERSTEIN. I would venture to say half a dozen times. I cannot quote an exact number.
Mr. HUNTER. Since when? Since the 1970's, or——
Page 71 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. Well, 1980.
Mr. WALLERSTEIN. I think the first supercomputer controls were instituted sometime in the early 1980's.
Mr. HUNTER. So 1980. So in 17 years we have changed it a half-dozen times. Let us accept that. You have given us the impression here that this is something you need to change on a very—on almost—at least the impression I got was this was something that could be adjusted weekly or monthly. But you have only changed it 6 times in 17 years. Why isn't 180 days for a fairly fundamental change reasonable?
Secretary REINSCH. Because you have to put it in the context of a process. This is not an instantaneous action. We don't send the memo to the President and 10 minutes later get an answer.
This administration has tried to base the policy changes it makes in this area on research, and that, we have gathered for us both internally within agencies and by independent consultants. That gathering an analytical process generally takes at least a year.
So what you have is a process that is more or less continuous; that is, for example, at the present process, because there is one ongoing, we let the contract for independent work in this area, I believe in February 1997, and that report, whatever it says, will be concluded probably January 1998. So it is an 11-month report. The President, the interagency process will then operate, and we will review that report. We will review our own work and I would anticipate the President would get a result probably in March, say. He will do whatever it is he has to do to it and deliver an opinion.
Page 72 PREV PAGE TOP OF DOC Segment 2 Of 2
It is already 2 1/2 years since the last time the President made a decision. What you are doing within a 6-month waiting period is making it more than 3 years, should he decide to make a change. And in our judgment, that is too long. It is not—your 180 days should not be viewed in isolation from the entire process.
Mr. HUNTER. OK, but at least the administration thinks it is important enough that they have a fairly lengthy review process, right?
Secretary REINSCH. Yes. This is a difficult decision.
Mr. HUNTER. So I—discharging our congressional duties, why should we not be as interested in at least as thorough a review?
Secretary REINSCH. I don't think it should take you as long to read the report as it took us to write it.
Mr. HUNTER. Should we accept everything you have put in the report as gospel?
Secretary REINSCH. And you usually don't, but I don't think it takes you that long a period of time to make up your mind.
Mr. HUNTER. I think you are right. Last question, Mr. Reinsch. Some of the people I have talked to in the nuclear weapons—in our nuclear weapons program here in the United States said two things, first, that the sales that went to the Soviet Union probably only very marginally aided their weapons programs, but that the sales of supercomputers that went to China substantially aided their nuclear development programs.
Page 73 PREV PAGE TOP OF DOC Segment 2 Of 2
Do you have an opinion on that? Do you agree or disagree?
Secretary REINSCH. I defer to the Department of Energy on that. I believe they have told you what they think.
Mr. HUNTER. Well, OK, so you—Mr. Wallerstein.
Secretary REINSCH. I won't contest their judgment.
Mr. HUNTER. Do you have an opinion on that?
Mr. WALLERSTEIN. We have opposed uniformly any assistance to those weapons programs and if there is evidence that the machines were used in those programs that would be detrimental to U.S. national security interests. But we have no specific information on how or to what extent they were used in either of those cases.
Mr. HUNTER. OK. I think that is—I think you gentlemen have made my point. My point is that at least you have some body of opinion in our nuclear weapons programs that this was bad for our country, adverse to American interests, and helped the Chinese substantially in their nuclear development programs. Neither one of you has a real—from what I understand, a real opinion on that or know too much about it.
Secretary REINSCH. I would not go as far as you did in your last statement. You snuck an extra little clause in there.
Page 74 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. Well, restate it for me, then.
Secretary REINSCH. I do not think—as I said, I am happy to consult with the Department of Energy, but I do not think that anybody has presented evidence that the computers you are referring to in China, whether it is 47 or the 2 under investigation, have actually been used for nuclear weapons development purposes. I think what you have been given is an assessment of what the impact would be if they were, which is a different thing.
Mr. HUNTER. Well, have the computers been retrieved?
Secretary REINSCH. The Sun Microsystems computer system has been retrieved. It reentered the country on November 9.
Mr. HUNTER. How about the others?
Secretary REINSCH. Well, there is only one other one that is under investigation in China right now, and it has not been. But I do not believe that—I am confident no one has presented any evidence that that one, or any of the others, for that matter, is being used for nuclear weapons research.
Your point earlier was that they could be diverted for that purpose, and I think that is a legitimate discussion. But I would not make at this point the leap that they have been.
Page 75 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. And you feel that you have a reliable fix on the location of all the computers at this point?
Secretary REINSCH. We are not——
Mr. HUNTER. Because we started out saying, you guys didn't know—you didn't know what the number was. You were going to get this information and get back with us.
Secretary REINSCH. We are not completely satisfied across the board, and we continue to investigate. We have not yet seen—if you want to be specific about China, we have not yet seen anything in China that has caused us to open an additional investigation, but we continue to——
Mr. HUNTER. Answer my question. I want to know what you have seen. Do you have a good, reliable fix on the other—the location of all the supercomputers?
Secretary REINSCH. With respect to some of them, yes, with respect to all of them, not yet.
Mr. HUNTER. OK. I think that is another point in favor of the chairman and the ranking member's bill or the conference bill. But, gentlemen and ma'am, thank you very much. I think it has been a good, worthwhile discussion. Appreciate it.
Thank you, Mr. Chairman.
Page 76 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. I think we have about explored this thing from this side about all we can, and we appreciate all your input. We will let you go and as somebody said we will hear the other side of the story now from our next panel. So, again, we appreciate it and hope your voice improves, Mr. Secretary.
Secretary REINSCH. Thank you, Mr. Chairman, it is always a pleasure to be here. I mean that sincerely. I wish I could stay for the second panel and then refute them, but I am sure I will have another opportunity.
Mr. HUNTER. Will you go for one out of two? We are with you on encryption.
Secretary REINSCH. Well, I appreciate that.
The CHAIRMAN. Gentlemen, we understand you have a tight schedule and we are going to try to hurry through. We appreciate your taking time to come and help us today with this problem, and we will just launch right on off.
As I said before, you can submit your prepared statements if you have them for the record, and we will proceed as you would like, starting off with Mr. Milhollin.
STATEMENT OF GARY MILHOLLIN, EXECUTIVE DIRECTOR, WISCONSIN PROJECT ON NUCLEAR ARMS CONTROL
Page 77 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. MILHOLLIN. Thank you very much, Mr. Chairman, my statement has been submitted and I ask the committee to incorporate it, in total, in the record.
The CHAIRMAN. Yes, granted.
Mr. MILHOLLIN. I would like to go back over a little bit of the ground I covered earlier with the committee, but very rapidly. I should say, the Subcommittee on Military Procurement.
I pointed out earlier this year that there had been a very dangerous sale by Silicon Graphics. We have discussed that today. Since that sale has been publicized, we have also learned about the IBM sale to Arzamas 16. We are now faced with a situation that the Russians are able to design weapons of mass destruction at two different sites with supercomputers imported illegally from the United States.
In my opinion, that was an inevitable result of the mistaken policy the administration has been following on supercomputer exports. These were disasters waiting to happen.
Also, in April, I pointed out that the Chinese Academy of Sciences had received an even more powerful supercomputer. Today, Mr. Reinsch admitted that it was about 5 billion operations per second; I put it at about 6, but in any case, it is a vast improvement in the ability of the Chinese Academy of Sciences to do military work, missile work and nuclear work. It does all of those things, has for a long time. In fact, it has the distinction of working on one of the few ICBM's China has that can strike American cities.
Page 78 PREV PAGE TOP OF DOC Segment 2 Of 2
I think we have to assume that since the Government does not know what this computer is being used for, that it can be accessed to make those missiles more accurate and to make the warheads they carry lighter and more powerful. I see that as a giant setback for U.S. national security.
I also think that the committee should be aware that these sales were not terribly ambiguous; that is, the law was clear at the time these sales were made. To sell a computer operating at the speeds of these computers required a license unless the computer was going to a site that was not military and not linked to weapons of mass destruction. In none of these cases was that true. The burden was on the exporter to determine that the exception to the license provision was available.
There seems to be some confusion at this time in the Government over what the law is and who has the burden of proof. I say that because these investigations do not seem to be going anywhere. The term under investigation is beginning to mean under the rug, in my opinion. We have been investigating the Silicon Graphics case for 9 months. The case is a clear case.
I think the committee should demand to be briefed on the status of the investigation of the Silicon Graphics export to Chelyabinsk on the status of the IBM export to Arzamas, on the Silicon Graphics export to the Chinese Academy of Sciences, and also the committee should be briefed, ask to be briefed, on two other diversion cases which have also been under consideration for some time. I am thinking now of the McDonnell Douglas case, which the committee is quite familiar with; that case has been around since 1994, and we still do not have a disposition.
Page 79 PREV PAGE TOP OF DOC Segment 2 Of 2
As Mr. Reinsch mentioned today, the Sun Microsystems diversion case is also under investigation. But I am afraid that unless this committee stays on top of this issue and demands periodic briefings, these investigations are simply going to languish and they are not going to produce any results.
The whole export control system is based on the idea that if somebody breaks the law, it is enforced. If we have a Commerce Department that is putting out the message that, first of all, we do not want to control this stuff, and second, when you break the law, we do not really want to investigate it, then the exporters are put—as Mr. Hunter has stated, they are put in a terrible position. They are told, do not break the law and then they are told, even if you do break it, something bad might happen to you or might not.
If I were an exporter I would not want to be in that position. And certainly, as an American citizen, I would not want to be in a position to think that my Government was not really interested in enforcing the law.
So I recommend to the committee that it be diligent in demanding that it be briefed. I think there needs to be some congressional oversight concerning these investigations.
With respect to the amendment, I would like to make a few points that were not made, even though it has been admirably discussed. First, I think it is actually a benefit to companies. The U.S. Government is giving you a bureaucrat who can tell you whether your—maybe I should not use that word—it is giving you an official who can tell you whether your export is a risk; I should think any company would want to take advantage of that. A 10-day wait to be told that it is OK, and that you are safe, seems to me a very good deal.
Page 80 PREV PAGE TOP OF DOC Segment 2 Of 2
Also, in exchange for this 10-day period, the Government can actually track which companies in the world are buying our computers. That is a valuable piece of intelligence that the Government would be getting for very little burden.
Finally, what would the administrative burden be? We have heard testimony that during the period when records were kept, about, I think, a 14-month period, only 91 cases or 91 exports went to tier III countries, 91. I did that basic arithmetic. That is an average of only two per week. So we are talking about two inquiries per week to the Commerce Department.
How much of an administrative burden could that be to handle two inquiries per week?
It certainly seems to me worthwhile to handle two telephone calls a week in order to keep computers out of the wrong hands. And given the CIA's budget, I am sure they could manage to field two inquiries per week, using their classified or unclassified databases. I do not think that is too much to ask in exchange for their recently declassified budget.
Finally, I think the committee ought to consider legislative remedies in the event that judicial remedies are not forthcoming. And as I said, they do not seem to be forthcoming at this time.
I recommend a principle to the committee—oh, by the way, I should say I am very proud to have recommended to the committee the amendment that the committee wound up recommending or wound up enacting. I am very proud of having done that. I recommended that the rollback in export controls be done in my testimony before the subcommittee, and I am glad I recommended that and I am glad the committee concurred with it.
Page 81 PREV PAGE TOP OF DOC Segment 2 Of 2
I would like to now make another recommendation. I would like to suggest a principle to the committee, and the principle is that if an American company is willing to illegally participate in a foreign nuclear weapons program or create the risk that its product could be used in a foreign nuclear weapons program, that it be barred from participating in our nuclear weapons program. If a company is going to help the Russians or the Chinese maintain their nuclear arsenal through illegal sales, I think that company should be barred from any participation in our stockpile program.
I think that would be a remedy that would be far more effective than these prosecutions that are going on, because the risk to the company would be far greater. The company would have to make a choice. Either it is going to accept the many millions of dollars that would come from participating in our program or it can take the nickels and dimes it gets from selling to marginal countries.
It is too bad Congresswoman Harman is not here. I was going to make the point that what we see from the data is that the sales to the dangerous countries are insignificant. Six percent of the sales that Mr. Reinsch recounted actually went to dangerous countries. Now, no computer company is going to live or die according to whether it makes sales to these countries or not, because it is such an infinitesimal part of the market. If a supercomputer company cannot succeed by selling to its main market, it will not succeed by selling to these marginal markets, where, in fact, most of the demand for high-speed computing is from the military. We are not talking about countries that have big civilian infrastructures that need a lot of high-speed computing; we are talking about countries that do not have these kind of infrastructures and that needs computing for military purposes.
Page 82 PREV PAGE TOP OF DOC Segment 2 Of 2
I think one last point I would like to make is the argument that you could send somebody here and set up a little shop somewhere and do weapons calculations and ship the results back. I think that would be a very dangerous proposition that no country would undertake. If we were to intercept a communication from a weapons design group operating for a foreign country in the United States, can you imagine what the reaction of this committee would be to that kind of a disclosure? I just do not think any foreign country would take such a risk. I just do not think it is in the cards.
Then there is the final point that, well, this is not controllable. I think that defies the evidence we see and what is actually going on. If it were possible to send somebody to the United States to buy a pocketful of computer chips and a soldering gun and then go back to Russia and make a supercomputer, the Russians would not have gone to the extraordinary lengths they did to smuggle in the supercomputers they just bought from us. So all the evidence we have indicates that the Commerce Department's analysis of what is controllable or not controllable is false.
I will stop there.
The CHAIRMAN. Thank you.
[The prepared statement of Mr. Milhollin can be found in the appendix on page 55.]
The CHAIRMAN. Mr. Bryen.
Page 83 PREV PAGE TOP OF DOC Segment 2 Of 2
STATEMENT OF STEPHEN BRYEN, FORMER DIRECTOR, DEFENSE TECHNOLOGY SECURITY ADMINISTRATION
Mr. BRYEN. Mr. Chairman, thank you very much for inviting me here today, and I would like to congratulate the committee for its fine work in this area. I am a strong believer in the importance of technology security as a policy and as a practice. We have gotten away from that, and I think the committee's amendment and its other actions will help redress the balance. So I commend you and Mr. Dellums, who has left, and certainly Congressman Hunter. I know it has taken a good deal of effort and courage to oppose the tide of events, but this is a case where I think it is more than justified.
My great concern is China. You can sweep all this other stuff away, as far as I'm concerned. China, I believe, is a potential adversary of the United States. I believe she is modernizing her military forces and particularly her nuclear forces to our detriment.
In my prepared remarks, which I ask you to submit for the record, please, I point out that China is developing, among other things, long-range cruise missiles, nuclear cruise missiles, which require rather small warheads to be successful and to have the range that they would like to have. We are the target of those; we are the primary target of those.
She is also acquiring other technologies. It is not just getting the supercomputers to achieve this goal; she is acquiring lots of other technologies that have been released through the policies of the current administration. To name a few of them: The global positioning system technology, which has been sold to China; some advanced engine technology which has been sold to China. So this is a problem that is bigger than just the supercomputer.
Page 84 PREV PAGE TOP OF DOC Segment 2 Of 2
But the supercomputer is a very crucial element, and the reason it is is that it will allow the Chinese weapons designers to build these small weapons and simulate the testing of these weapons. This is a nontrivial exercise. It is extraordinarily important, and it will take them a few years to learn these efficiently, but I have little doubt that they will do so.
Of the total number of so-called tier III sales that have been made in China to key beneficiaries, overwhelmingly so, 47 of 91 machines have been sent to China; and in particular, as Gary has mentioned, to the Academy of Sciences, the Chinese Academy of Sciences, which is a primary nuclear weapons design center. I have been there; I have seen their equipment in the late 1980's. It is pretty impressive, and clearly, it is a high priority of the Chinese Government to see to it that that center functions in designing and working on the design of these advanced weapons programs.
The excuses we have heard today for the program, as far as I am concerned, are totally irrelevant. If you just take China and sit it to the one side and say, we have to deal with this problem because it is a potential threat to our survival, then we can surely come up with a policy that will work.
On the other hand, if we want to feed China with this kind of technology, then we surely have come up with a policy that does not work. And that is putting it in the most simple way.
I would urge the committee to take a very close look at what kind of technology transfers are going forward with respect to China and what the long-term impact will be on us, because it is happening at a time when we are pulling in our horns as far as our defense programs are concerned.
Page 85 PREV PAGE TOP OF DOC Segment 2 Of 2
The slashing and burning of the defense budgets is something I am sure you are quite intimately familiar with. It worries me a great deal that China will be able to counter our freedom of action in the Pacific area, and that has long-term negative consequences for the United States, for United States security, and by the way, for our economic security as well.
In my prepared remarks, I went into an area that is somewhat different, but I wanted to raise it today because it is of great concern to me. There is a reorganization going on in the Defense Department. It has been in the newspapers. One of the victims of that reorganization has been—is the proposed transfer of the Defense Technology Security Administration out of the policy cluster, and the elimination of the Joint Chiefs of Staff having the capability to review license applications.
There has been a lot of talk about Secretary Cohen looking at the fruits of your amendment, but the agency that is supposed to do that will no longer be in the policy cluster and will no longer play a role in this nor will the JCS be informed.
It is a very negative policy. It is a very bad idea to downgrade and relegate to a marginal role this critical agency. I was the founder of the agency, so I think I know what I am taking about here; and I would really urge the committee—I know Secretary Cohen has the right to organize his Department as he sees fit, but I would really urge the committee to hold a separate hearing on this issue to see if this cannot be turned around, because it is a mistake.
For the purposes of controlling proliferation, you will have a proliferation shop in the policy cluster with no capability to do any analysis. You will have a Joint Chiefs of Staff that will know nothing about anything, because it will not be getting fed the information on what is passing through the licensing mill that is so crucial. And licenses are one of the ways you can determine and understand what is happening and what is going abroad and where it might end up.
Page 86 PREV PAGE TOP OF DOC Segment 2 Of 2
So if it pleases the committee, I hope you will have a look at this and try and persuade the Secretary that this is a very bad idea, it is a dangerous idea, and it is one that will not serve him or anyone else in this country well.
Those are my main points, Mr. Chairman, and I will be glad to take questions.
[The prepared statement of Mr. Bryen can be found in the appendix on page 65.]
The CHAIRMAN. Mr. Skelton.
Mr. SKELTON. No questions, Mr. Chairman.
The CHAIRMAN. Mr. Hunter.
Mr. HUNTER. Thank you, Mr. Chairman. I just want to thank our two experts. It was largely, I think, a function of your concern and your wisdom that we injected into this process that helped to percolate up and build the package that the chairman and the ranking member have put together.
It kind of goes to the heart of the illogic of the administration's position that they do not want people exercising judgment when judgment is everything in this area. It is just difficult to understand what their basic thinking is. And what you say, I think, is very clear.
Page 87 PREV PAGE TOP OF DOC Segment 2 Of 2
One thing, Mr. Milhollin, the one point you made that we always need to remember is, when people want to loosen these things up, they always come to us and tell us the cat is out of the bag, you cannot put the genie back in the bottle, everybody has them; and that export control, for practical purposes, is a disappearing science, or it is something that has less and less relevance as time marches on. And yet when you made that post-Desert Storm analysis that showed us that, in fact, export control in many areas had worked, had kept Saddam Hussein from having as much as he could have had in terms of capability at that time, I thought that was a very compelling argument for doing everything that we can do to stem the tide of this technology.
This is so absolutely critical. The idea that we fight, we hold up the defense bill for weeks and weeks because of political battles that take place in the administration over things like base closing, and yet here we want 10 days for the Secretary of Defense to look at a supercomputer transfer that could be used to make a nuclear weapon. We just do not have time for that and we just do not have time to take those two calls per week.
I think you have said everything that needs to be said, and I take it that you fully and totally endorse the conference package that we have here; is that accurate?
Mr. MILHOLLIN. I certainly do. I think it is a tremendously beneficial step in the right direction, both for companies and for the Government and for security.
I guess I would have preferred for the language that the committee originally came up with to have prevailed, but since it did not, I must say that I think most of what was most important survived and that the committee ought to be proud of its work. I fully agree that this is something that is needed, and I hope the President signs it.
Page 88 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. HUNTER. Thank you.
Mr. BRYEN. I agree. I think it is a real step forward, as I said in my opening remarks. This is a real achievement that the committee has made in the interest of security, and I hope it prevails.
Mr. HUNTER. I want to thank our leader, the chairman of the committee, for being tough on this very important area. We have to arm-wrestle with the administration to get this darned bill signed. We are not just a money committee. This is one of those other areas where we cannot afford for national security purposes to let go, just to move things quicker through the pipeline.
So thank you, Mr. Chairman.
The CHAIRMAN. Thank you. And again I want to thank you gentlemen. Your expertise in these matters has helped us a lot already, and we certainly appreciate it. I wish everybody in this country could see what is happening. This is an example of the frustration we have in trying to provide for our national defense in this country.
Mr. SKELTON. Mr. Chairman, I just want to say thank you to them, as well, for their expertise. I heard every word you said and we agree.
Mr. MILHOLLIN. Appreciate it.
Page 89 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. BRYEN. Thank you.
The CHAIRMAN. I appreciate the input from the gentlemen.
As I was saying, this is just one example, another example, of the problems we face in trying to properly defend this country. I guess we are programmed to do battle with opponents, external opponents, threats to our country, and we have no experience in doing battle with our own administration, which includes all the other agencies of the Government controlled by the administration.
I was just sitting here talking to staff about a long litany of things going from national missile defense, which we try to provide, and even got the CIA to issue a politicized intelligence estimate downplaying the threat, saying except for declared nuclear powers, we do not face any threats for 10 or 15 years.
We try to provide a theater missile defense. Everybody says that is one of our biggest priorities. We even got the architecture provided. It was signed into law, the bill. And then they refused DOD, and the President refused to go along with the law. We had to bring a lawsuit against our own Government to forcefully go along with it.
We issued a big report on readiness of our military which they denied and then tried to put in $20 billion to show that they were, they did not have a problem. I can go on and on and on, right on down the list, and it is just frustrating to try to—that every time we try to do something in Congress under article I, section 8 of the Constitution—that is our power and duty to provide for our military—the executive branch administers what we give them. But to tell us what they are going to administer is something different, and it is just a frustrating experience. And I apologize for burdening you with these problems we have.
Page 90 PREV PAGE TOP OF DOC Segment 2 Of 2
Mr. BRYEN. It is not a burden, Mr. Chairman.
The CHAIRMAN. It is another example of what we are talking about here today, and this is another thing I can add to the list, a line-item veto I just might throw into the pot, too.
Mr. BRYEN. You have to keep fighting for those programs, Mr. Chairman, because they are very important to this country.
The CHAIRMAN. We have to let the American people know just what is happening, and that is what we are trying do. We appreciate your effort, and we apologize for keeping you so long. I understand you have to make some connections. Again, thank you very much for your contribution.
Mr. BRYEN. Thank you.
Mr. MILHOLLIN. Thank you, Mr. Chairman.
[Whereupon, at 1:12 p.m., the committee was adjourned.]
"The Official Committee record contains additional material here."
QUESTIONS SUBMITTED BY MR. SPENCE
Page 91 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. In your testimony on November 13, 1997, you mentioned that a U.S. supercomputer inappropriately diverted to China's Changsha Institute of Science and Technology was returned to the United States on November 9, 1997. Why did it take so long to obtain the return of this machine? Was there a quid pro quo? Do we know how that computer was being used? What is the present location and the status of the computer? Has it been returned to the original exporter?
Secretary REINSCH. The computer at issue was the subject of a transaction in which governments and private parties in both the United States and China had an interest. The time taken to return the computer does not appear unreasonable in light of the sensitivity of the multiple interests involved. The United States Government made no agreements or promises to anyone to get the computer returned. Because this transaction relates to a matter that is under investigation, it would not be appropriate for me to discuss how the computer was being used. The computer is currently in the United States in the possession of the original exporter.
The CHAIRMAN. Is the United States seeking the return of any other unlicensed supercomputer transfers to China?
Secretary REINSCH. Not at this time.
The CHAIRMAN. Have any additional unlicensed supercomputer transfers to China or Russia taken place or been discovered in the past six months? To any other Tier III countries?
Secretary REINSCH. We have opened one additional investigation with regard to a computer export to China and continue to review other sales as part of the regular export enforcement process.
Page 92 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. There have been press reports indicating that a supercomputer illicitly shipped to a Russian nuclear weapons research laboratory is ''missing'' and may have been diverted to Iran. Do you have any information to substantiate these reports? Could such a diversion have occurred? How confident are you that the U.S. government would know about it?
Secretary REINSCH. The Russians have indicated to us that the computer in question is in storage and has not been delivered to its intended end user. We have not yet been able to verify that statement. We have no information that would substantiate the press reports to which you refer. The only way that we can be absolutely certain that U.S. origin items are not diverted would be to prohibit all exports. I believe that with all of the resources available to the U.S. Government, it is unlikely that such a diversion would take place without us learning about it.
The CHAIRMAN. An October 4, 1997, New York Times article stated that Libya had acquired U.S.-origin computers through front companies operating in Europe. What kinds of U.S. computers has Libya been able to acquire and for what purposes are they being used? What is the computational capability of these computers? Is the United States attempting to have them returned?
Secretary REINSCH. The article to which you refer notes that the case it discusses is under investigation under the leadership of the Department of Justice. We investigate many allegations of diversions to Libya. It would not be appropriate for me to discuss any specific matter that is under investigation.
Page 93 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. Is there any procedure in place to allow the export of U.S.-origin items under the terms of a classified export license? Have any supercomputers been transferred to any destination under the terms of such a classified license? If so, please provide details regarding how many have been transferred and when, to what end-users, for what stated end-use, and their computational capabilities. If such transfers have occurred, are they included or excluded from the previously reported total of U.S.-origin high-performance computers shipped to other countries?
Secretary REINSCH. No exports of high performance computers have been authorized under a classified export license.
The CHAIRMAN. Has the United States sought any guarantees from Russia or China that U.S.-origin supercomputers will not be used in weapons programs? If so, what has been the response? If not, why not? Have we sought such guarantees from any other Tier III countries?
Secretary REINSCH. Since the January 1996 implementation of revised computer export controls, the U.S. has not approved any individual validated licenses (IVL) for high performance computers for export to Russia or the People's Republic of China (PRC). Therefore, it has not been necessary for the U.S. to seek guarantees from Russia or the PRC in recent years.
Prior to 1996, one license was approved in May 1995 for an HPC computer export to Russia. The United States sought and received a government certified safeguard plan that the computer would be used as authorized. This system would now be eligible for export without a license using License Exception CTP, as it was for export to a civil end-user. Because of the end-user and the CTP eligibility, the safeguard plan is now not in place. Prior to 1996, some exports were approved for HPCs to China with safeguard plans certified by the Chinese government, but these were for computers at relatively low CTP levels.
Page 94 PREV PAGE TOP OF DOC Segment 2 Of 2
We have sought and received such guarantees in the form of safeguard plans from other Tier III countries, such as Israel and Saudi Arabia, in recent years. All safeguard plans are approved by all relevant agencies.
The CHAIRMAN. In a November 10, 1997 interview with Defense Week, you stated that Russia and China ''are not going to use them [U.S. supercomputers] to build weapons.'' How can you make such an unequivocal statement regarding how Russia or China will use U.S. supercomputers, especially since the United States has not been allowed to conduct any post-shipment verifications of the computers that were inappropriately shipped to weapons-related facilities?
Secretary REINSCH. My response was in regard to an allegation by Defense Weekly that Russia and China were ''going to actually use them [U.S. supercomputers] to build weapons.'' The purpose to which the Russians allegedly put the computers was weapons stockpile maintenance. What we have not approved HPC exports for that purpose, it is not the same as building weapons. The allegation in the Chinese case was that the computer went to an organization engaged in military research. Again, this is not the same as building nuclear weapons. I made this distinction in my answer to counter the implication of the question that the end use of these computers was more detrimental to our national security interests than have been previously alleged.
The CHAIRMAN. In your November 13, 1997 testimony you stated that 91 supercomputer were exported to Tier III countries between January 1996 and March 1997. However, you also stated, ''Thus far in 1997, BXA has requested 22 post-shipment verifications of [high-performance computers] exported to Tier III countries.'' How many post-shipment verifications (PSVs) were requested by BXA in 1996? How many were conducted? Do the 22 PSV requests in 1997 involve different computer exports, or have there been multiple requests to conduct verifications on the same export? Why were no PSVs requested for the other exports? To what countries were these computers exported? Please identify those computer exports where no PSVs were requested as a result of regular service and maintenance performed by the exporter.
Page 95 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. BXA neither requested nor completed any post shipment verifications of high performance computers shipped to Tier III countries in 1996. Subsequent requests were each for a different computer export. A decision to request a PSV is made based on criteria BXA has developed. Post shipment verifications were not requested for other computer exports because the end users identified on the license application met criteria which, in BXA's judgement, indicated no post shipment checks were necessary.
Post shipment verifications were not requested for exports to China because China does not currently allow the Embassy to conduct these checks. The Tier III countries for which PSVs were requested were Russia, Saudi Arabia and Israel. More public discussion of specific cases is precluded by Section 12(c) of the Export Administration Act.
The CHAIRMAN. In a July 22, 1997 letter to Chairman Spence on the issue of supercomputer exports, the President stated that ''my Administration has adequate administrative and enforcement means under current law to address problems that arise with U.S. computer exports.'' Do you believe that these means are appropriate to deal with the refusal of China and Russia to return U.S. supercomputers that have been shipped without proper authorization?
Secretary REINSCH. Yes, I do. The fact that the computer sent to China has been returned is evidence that we have effective means to address problems. In addition, we possess a wide range of policy and enforcement options to deal with violations of our regulations. We are committed to using all of the tools at our disposal to implement our export control policy and punish violators.
Page 96 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. What legal action is being taken against U.S. exporters who wrongfully shipped supercomputers without export licenses?
Secretary REINSCH. We are investigating these allegations. As I have stated, it would not be appropriate for me to comment on matters under investigation. We are committed to conducting thorough investigations and imposing appropriate criminal and administrative sanctions on anyone who violated our regulations.
The CHAIRMAN. The President's National Security Advisor, Samuel Berger, in an October 20, 1997 letter to Chairman Spence, stated ''we are making progress regarding the unauthorized computers shipped to Russian nuclear labs.'' What specifically is the nature of this ''progress''? Has the United States asked Russia to return these computers? If so, what has been the Russian response? If not, why not? Does the Administration plan to request from Russia the return of the 16 additional high-performance computers shipped to Arzamas-16 through European intermediaries?
Secretary REINSCH. We continue to engage the Russian government at senior levels concerning the return of the computers in question. The Russians have informally indicated that they may be considering transferring the computers to an unclassified location in Russia, but we have not yet received a formal proposal to this end. We continue to pursue a dialogue with the Russians on this issue. In addition, the Department of Justice has made a request for cooperation under our Mutual Legal Assistance Agreement and is continuing to work on the issue through legal channels. It should also be pointed out that the ''16 additional high performance computers'' are neither additional nor high performance. None of them is a high performance computer within the parameters of the President's policy (i.e. over 2,000 MTOPS).
Page 97 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. In your November 13, 1997 testimony you stated that ''Russian officials have indicated that Russia is considering moving the computers in question to facilities where unclassified research is being done.'' To which computers were you referring? Has Russia moved these machines yet? What are the names of those facilities and where are they located? Will Russia allow us to conduct verifications of their whereabouts and use at these unclassified facilities? If not, what difference does it make where the Russians move these machines?
Secretary REINSCH. See answer to question #12. Discussions on these questions are ongoing. It would be premature for me to comment further.
The CHAIRMAN. If Russia refuses to return the U.S.-origin computers that have been transferred to Russian nuclear laboratories without proper authorization, will it allow the United States to conduct post-shipment verifications of their location and use?
Secretary REINSCH. That is one of the matters that we are discussing with the Russians.
The CHAIRMAN. Did the Administration ever suggest to Russian officials that the acquisition of U.S. supercomputers could help Russia maintain the reliability of its nuclear weapons stockpile in the absence of nuclear testing? Did Russia ever indicate that its agreement to the Comprehensive Test Ban Treaty would be conditioned on access to U.S.-origin supercomputers? If so, what was the Administration's response?
Page 98 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. The Administration did not suggest to Russian officials that the acquisition of U.S. supercomputers could help Russia maintain the reliability of its nuclear stockpile in the absence of nuclear testing, nor did Russia indicate that its agreement to the CTBT would be conditioned on access to U.S.-origin high performance computers.
The CHAIRMAN. Beyond any changes in policy that may result from this year's National Defense Authorization Act, what future actions does the Administration contemplate taking to ensure that U.S. supercomputers are not exported inappropriately?
Secretary REINSCH. We are working with U.S. industry to ensure that it and its overseas subsidiaries understand their responsibilities through an aggressive outreach program. If exporters are unsure of the activities of potential end users for their computers, they are urged to seek guidance from BXA. We will also publish the names of additional end users of concern which will not be eligible for shipments under license exception CTP. We will also continue to require quarterly reports on the end users and end uses of HPCs which have already been shipped. Should ongoing or future investigations lead to either criminal or civil penalties, those penalties will act as a deterrent to other exporters.
The CHAIRMAN. The November 10, 1997 issue of Defense Week reports that a ''special comprehensive license'' was created in March 1996 that would allow ''high-volume exporters to export most types of equipment without obtaining specific licenses for each country.'' How does this ''special comprehensive license'' differ from a ''general destination license''? How many ''special comprehensive licenses'' have been issued since March 1996? What specific items have been exported under this license since March 1996, to what destinations, and for what end-use?
Page 99 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. The most significant difference between a special comprehensive license (SCL) and a ''general destination license'' (i.e., license exception or authorization to export without an export license) is that an SCL requires prior authorization from the government before items may be exported and a license exception does not. Six SCLs have been issued since March 1996. As a condition of approval the exporter and consignee must implement an Internal Control Program (ICP) which is a system of checks to ensure compliance with the Export Administration Regulations and the conditions of the license. These ICPs are reviewed and approved by BXA.
A License Exception is not a license, but an authorization available to exporters when their transactions fall within specifically defined parameters to export without prior BXA approval or documentation. A ''Special Comprehensive License'' (SCL) is a license type that was created for knowledgable, experienced exporters who ship frequent, high volume items to BXA pre-approved consignees. It is designed to reduce the paperwork burden on exporters by eliminating the need to submit so many individual applications.
Since March 1996, eight company-held SCLs and one U.S. Government agency-held SCL have been issued. Four had computers approved. Three of the four licenses authorize computer exports, but not above license exception CTP limits. The fourth involved computers well below high performance computer levels. SCLs are reviewed by all relevant agencies, except where agencies have granted a Delegation of Authority (DOA) to the Department of Commerce for certain items, and in certain instances where an exporter already had prior approval under a Distribution License (the forerunner to the SCL) for such exports.
Page 100 PREV PAGE TOP OF DOC Segment 2 Of 2
The CHAIRMAN. Is the Commerce Department regularly updating the supercomputer export list with information provided by the exporters, or is that information only being requested by the Commerce Department on an as-needed basis?
Secretary REINSCH. We have been requesting the data on a quarterly basis.
The CHAIRMAN. When will the latest study on supercomputer export controls, conducted by Stanford University under contract to the Departments of Commerce and Defense, be complete? When will it be forwarded to the Department of Commerce and the President? And how soon after that might the President be expected to review the applicability of existing export control thresholds and make recommendations for change, if appropriate? Please provide the committee with a copy of the Request for Proposal (RFP) for this study, copies of the responses received, and a copy of the contract which resulted from the contract award.
Secretary REINSCH. Agencies are being briefed on the latest study on export controls for high performance computers in January, 1998. Subsequently, the agencies will forward recommendations to the President for his consideration. Attached are copies of the RFP, the sole proposal submitted in response to the RFP, and the contract.
[The information referred to can be found on p. 79.]
QUESTIONS SUBMITTED BY MR. GIBBONS
Mr. GIBBONS. What was the MTOPS speed of that computer?
Page 101 PREV PAGE TOP OF DOC Segment 2 Of 2
Secretary REINSCH. I promised to provide the exact MTOPS level for the Silicon Graphics computer shipped to the Chinese Academy of Sciences. That number is 5,908.69 MTOPS and should be inserted [in] the hearing transcript.''
"The Official Committee record contains additional material here."