[Page: S6877]

[Page: S6894]

Mr. COCHRAN addressed the Chair.

The PRESIDING OFFICER. The Senator from Mississippi.


Mr. COCHRAN. Madam President, I know I don't need to ask consent to return to the Cochran amendment. But the Lugar amendment has been offered and has been the pending business. I ask that we return to the regular order, to amendment No. 420.

The PRESIDING OFFICER. The Senator has that right.

That is now the pending amendment.

Mr. COCHRAN. Madam President, amendment No. 420 was offered by me, and is cosponsored by the distinguished Senator from Illinois, Senator Durbin. It seeks to modify the existing export control policy that had been instituted by the administration with respect to the exporting of high-performance or so-called supercomputers.


Madam President, on November 14, 1994, President Clinton issued Executive Order 12938, the Emergency Regarding Weapons of Mass Destruction, declaring that the proliferation of weapons of mass destruction and the means of delivering them constitute `an unusual and extraordinary threat to the national security, foreign policy, and economy of the United States,' and that he had therefore decided to `declare a national emergency to deal with that threat.' The President reaffirmed this Executive order on November 15, 1995, and again on November 11, 1996.

We have had several hearings recently on the subject of proliferation in my Governmental Affairs Subcommittee on International Security, Proliferation, and Federal Services. And the distinguished ranking member of the full committee, Senator Levin, is the ranking member of that subcommittee.

We have examined cases of proliferation by the People's Republic of China and proliferation by Russia, and I can tell you that the facts--brought out in open session--are disturbing. The facts tell a story of both Chinese and Russian sales of technology, components, and delivery systems for weapons of mass destruction, as well as sales of highly capable advanced conventional weapons and other critical military technologies, to nations like Iran. The facts demonstrate that President Clinton was entirely correct in describing this problem as a national emergency.

Just last month, the Director of Central Intelligence sent Congress an unclassified report entitled, `The Acquisition of Technology Relating to Weapons of Mass Destruction and Advanced Conventional Munitions.' The report covers only the period July through December 1996 and levies serious proliferation charges against, among others, Russia and China. The report says:

China was the most significant supplier of WMD-related goods and technology to foreign countries. The Chinese provided a tremendous variety of assistance to both Iran's and Pakistan's ballistic missile programs. China also was the primary source of nuclear-related equipment and technology to Pakistan, and a key supplier to Iran during this reporting period. Iran also obtained considerable CW-related assistance from China in the form of production equipment and technology.

The intelligence community report--and I note that this report is not the product of any single part of the intelligence community, but represents the consensus view of the entire intelligence community--goes on to say, and again I quote:

Russia supplied a variety of ballistic missile-related goods to foreign countries during the reporting period, especially to Iran. Russia was an important source for nuclear programs in Iran and, to a lesser extent, India and Pakistan.

Madam President, the facts that emerged during my subcommittee's hearings on Russian and Chinese proliferation are completely supported by this latest report of the intelligence community. And we should not be comforted by the fact that it reports on the proliferant behavior of these nations only during the last half of 1996. For those who claim that Chinese and Russian behavior on proliferation is getting better, the best I can say is that it certainly is not yet good enough.

I raise the issue of proliferation because it is the principal reason we have offered this amendment on supercomputer export controls. The use of high-performance computers to upgrade existing weapons capabilities or develop new ones is not some fantasy or something that might happen in the future. It is known fact. High-performance computers help make it possible to develop and improve weapons capabilities that threaten the United States. Keeping them out of the wrong hands makes America safer. Dr. Seymour Goodman, in a report used by the administration as its basis for weakening U.S. export controls on high-performance computers, wrote:

. . . continued export controls will slow the exacerbation of existing nuclear threats. Control of HPC [high-performance computer] exports, by limiting those exports or imposing appropriate safeguards, to countries known to possess nuclear weapons will impede their development of improved weapons and reduce their confidence in their existing stockpile by limiting the opportunity to conduct simulations in lieu of live tests. Similar or more rigorous controls on HPC exports to countries with nuclear weapons development programs could impede their development of second-generation weapons.

The June 1997 Intelligence Community report to Congress couldn't be more clear on this issue. It states:

. . . countries of concern continued last year to acquire substantial amounts of WMD-related equipment, materials, and technology, as well as modern conventional weapons. China and Russia continued to be the primary suppliers, and are key to any future efforts to stem the flow of dual-use goods and modern weapons to countries to concern.

This amendment will help reduce the proliferation danger facing the United States by requiring an individual validated license to export all supercomputers to so-called Tier 3 countries, which include China and Russia. Because of the new export control policy for supercomputers announced by the Clinton administration on October 6, 1995, there currently is no such requirement. We must act to change that policy now.

This policy, which has been in place for almost 18 months, groups all nations into four country tiers and establishes export licensing requirements for high-performance computers based upon their country of destination. Tier 1 countries, consisting primarily of our NATO allies, are free to receive high-performance computers of unlimited capability without an export license from the United States, while, at the other end of the spectrum, Tier 4 countries, consisting of the last trustworthy, cannot legally receive any of these supercomputers. Almost all countries in South America, Central America, the Caribbean, and Africa are in Tier 2, and can receive supercomputers capable of up to 10,000 MTOPS--MTOPS are Millions of Theoretical Operations per Second, the standard measure of computing capability--before an export license is required.

The end-use and end-user are the critical factors for exports to any of the 50 nations comprising Tier 3. If the end-use and user are civilian, the policy allows exports of supercomputers capable of up to 7,000 MTOPS before an export license is required. If the Tier 3 end-use or user is military, U.S. export licenses are required for any high-performance computer capable of more than 2,000 MTOPS. But it is the U.S. exporter, not the administration, which has the responsibility under this policy for

determining the end-use and user for Tier 3 exports between 2,000 MTOPS and 7,000 MTOPS. This responsibility, difficult under any circumstances, is complicated by a company's natural focus on making sales. Our amendment addresses only these Tier 3 exports, as depicted by the diagonally-striped area on this chart, which I am going to show the Senate at some point in this discussion.

Our amendment applies to only a small portion of high-performance computer exports. In fact, according to the Commerce Department's Bureau of Export Administration, of the 1,436 supercomputers exported from the United States from the date the new policy went into effect through March 1997, only 91 went to Tier 3 countries. That amounts to 6.34 percent of total supercomputer exports. Does it not make sense for our Government to be willing to check to make sure that 6.34 percent of our supercomputer exports go to the right place? Is it unreasonable to require the administration to be sure that American supercomputer sales aren't going to people and places who would damage American national security?

Our amendment doesn't prohibit the transfer of a single supercomputer. It requires that the existing standards for transfers be monitored by our Government. Our amendment changes only one aspect of the policy, shifting the burden of determining end-use and end-user in Tier 3 countries from the exporter to the administration. Why is this so important? Listen to another part of last month's report to Congress by the Intelligence Community, which says, `Many Third World countries--with Iran being the most prominent example--are responding to Western counterproliferation efforts by relying more on legitimate commercial firms as procurement fronts and by developing more convoluted procurement networks.'

American exporters are not capable of determining whether a potential purchaser is a `procurement front' or part of a `more convoluted procurement network,' and it is wrong to place this burden on them.

The administration, and many exporters, will tell you that the current policy is working, that closer scrutiny isn't required, but look at this chart and what it shows you. There are American supercomputers in Russia's and China's nuclear weapons complexes. According to Russia's Minister of Atomic Energy, these supercomputers are `10 times faster than any previously available in Russia.' According to the Chinese Academy of Sciences--which works on everything from the D-5 ICBM, capable of reaching the United States, to uranium enrichment for nuclear weapons--its American supercomputer provides the Academy with `computational power previously unknown' and is available--this is a quote from them--to `all the major scientific and technological institutes across China.' American high performance computers are now available to help these countries improve their nuclear weapons and improve that which they are proliferating, courtesy of a policy that can be called many things, but can't reasonably labelled as `working.'

Just last week we learned through press reports that an American supercomputer sent to Hong Kong is now in China under the control of the People's Liberation Army. In addition to the 47 American supercomputers that have been shipped to China since this new policy took effect, 20 unlicensed American supercomputers have been shipped to Hong Kong. At least now we know where one of the Hong Kong supercomputers is. What about the others? does this look to anyone like a policy that's working? This is a real problem. It is a problem that exists now. It is not a hypothetical problem. It is not a problem that may develop in the future. This is a serious problem that threatens our national security.

There are some opposing this amendment who claim that setting the threshold at 2,000 MTOPS is too low, and consequently will make it impossible for American computer manufacturers to sell personal computers--PC's--abroad. That is just not true. It is a last minute desperation shot at the Cochran-Durbin amendment. Let's look at the facts:

The first fact is the 2,000 MTOPS threshold opponents express concern over was not dreamed up by us. It is the administration's limit.

No. 2, industry suggests that by some time in the fourth quarter of 1998--this date came, incidentally, from IBM's Director of public policy, who recently visited with my staff about this amendment--IBM will produce, according to him, a PC capable of just over 2,000 MTOPS for sale in the international marketplace, he said. But IBM couldn't answer several basic questions about this PC. Its Director of public policy didn't know the name of the PC, the expected price that would be charged for it, how many would be produced for the U.S. market, how many would be produced for potential foreign market sales, or even how many would be produced for this Tier 3 market, which this amendment is narrowly related to. It is worth remembering that this amendment that we are talking about only affects Tier 3 countries, and he's talking to us as if our amendment affects all sales to everybody--in the United States, foreign countries, everywhere--and that is just not true.

IBM doesn't just build these machines overnight on an impulse or a whim or a guess about what is out there in terms of potential sales. If it is going to have a new top-of-the-line PC out within 15 to 16 months, as they claim through this director of public policy, it must already have ordered the chip to run this PC. Doesn't it stand to reason that if such a PC were just around the corner, IBM would be able to answer some of these basic questions that I said the director could not answer? If not, is it possible that IBM is being overly optimistic about its capability, its projections, about the timeframe involved, and all the other arguments that have been advanced against this amendment?

Fact No. 3: Right now, according to William Reinsch, who is the Under Secretary of Commerce for Export Administration, `High end Pentium-based personal computers sold today at retail outlets perform at about 200 to 250 MTOPS.'

Did you hear that? We are not talking about 2,000 to 7,000 MTOPS, like some of these computer lobbyists are saying to Senators are going to be affected by this amendment. The PC's that are out on the market today are at much lower ranges of capability.

Let's give Secretary Reinsch the benefit of the doubt and say today's top-end PCs are capable of running at 250 MTOPS. Secretary Reinsch said on June 11 before my subcommittee in an open hearing that `computer power doubles every 18 months, and this has been the axiom in the industry for, I think, about 15 years.'

This axiom is known as Moore's Law. The math is straightforward. If top-end PC's are capable of 250 MTOPS today, 18 months from now they will be capable of 500 MTOPS; 36 months from now, they will be capable of 1,000 MTOPS; 54 months from now, in 4 1/2 years, they will be capable of 2,000 MTOPS. Fifty-four months from now is not, contrary to the claims of some computer manufacturers, the fourth quarter of next year, as was suggested to us by the director of public policy of IBM. Of course, Moore's Law doesn't even mean that 54 months from now there will be PC's on the market capable of 2,000 MTOPS. It only suggests that our manufacturers should be able to build these powerful PC's 54 months from now, if Moore's Law continues to be sustained. None of our manufacturers will build PC's this powerful unless there is a broad market demand for such a highly capable PC, and it is unclear if the market will even be demanding such a powerful PC many times more powerful than today's top-of-the-line PC's in just under 5 years.

If 4 or 5 years from now industry's optimism proves to be correct, I will be pleased to return to this floor and offer legislation modifying the 2,000 MTOPS level. But the suggestion that by next year we will have PC's many times more powerful than our most powerful today can only be guesswork, wishful thinking.

Fact No. 4: IBM currently sells, again according to its director of public policy, a workstation that is capable of just over 2,000 MTOPS. Wouldn't it make sense that future demand for the much anticipated 2,000 MTOPS PC should be similar to the current demand for the workstation that is already on the market?

According to the Commerce Department, from January 25, 1996, when the administration's supercomputer export control liberalization took effect, to March of 1997, 1,436 American high-performance computers were exported to countries in tiers 1, 2, and 3. Of these 1,436, just 91, or 6.34 percent, went to tier 3 countries. I do not know how many of these 91 were IBM's workstation that is just over 2,000 MTOPS. We know that at least 6 of the 91 were not manufactured by IBM--4 Silicon Graphics machines that are now running at Russia's nuclear weapons labs; one Silicon Graphics machine in the Chinese Academy of Sciences, which is a key part of China's nuclear weapons complex; and one Sun Microsystems machine that we just learned last week is now running at a Chinese military facility in Chungsha after being diverted from Hong Kong. So up to 85 of the 91 exported over 14 months to tier 3 countries could have been this IBM workstation, though I doubt that all of them consisted of that one machine. But even if all 85 were these IBM workstations, does this sound like the kind of volume that will overwhelm the Government's licensing apparatus? Certainly not.

The specter of American jobs being lost to unwieldy export controls is just another part of the argument against the Cochran-Durbin amendment that is not based on the facts.

Another argument made against our amendment is that the right way to keep organizations from getting American supercomputer technology who shouldn't be receiving it is for the Department of Commerce to publish a list of prohibited end users with individual validated licenses required for any high-performance computer export to a country or entity on the list. This argument against the amendment at least has the virtue of implicitly admitting that American supercomputers should not be in Russia's and China's nuclear weapons design labs, but it is another argument that is simply not based on the facts.

Shortly before the recent July Fourth recess, I spoke on

the floor of the Senate explaining why such a list would be, in many ways, worse than the current situation. I won't go through all those reasons again in the interest of time now, but I continue to believe that such a list would be necessarily incomplete because of the requirement to protect intelligence sources and methods. It could be used as the Department of Commerce's guide for proliferators, and it would make it only too easy to make a sale to a location not on the list, thus encouraging makers of weapons of mass destruction to establish phony front organizations for the purposes of acquiring U.S. supercomputers. They wouldn't be on that list.

In fact, the Department of Commerce on June 30 published such a list, and its inadequacy is obvious. The June 30 list, called by the Commerce Department the `Entities List,' consists of 13 locations in 5 tier 3 countries that can receive an American supercomputer only if you have a license, only subject to a license. So now the total list of proscribed end users consists of 15 entities. On this list are Chelyabinsk-70 and Arzamas-16 in Russia which have already received at least five American supercomputers and parts of the Chinese Academy of Sciences, which also is now manufacturing more modern nuclear weapons with America's finest technology.

Because of this list, now America's computer exporters know that they need a license to ship a high-performance computer to any of these entities. What about other entities, though? What about the Chinese company that shipped ring magnets to Pakistan last year for use in its nuclear program? Why isn't that company on the list? It has been subjected to sanctions imposed by our Government, and it is not on our Government's list as a prohibited end user. What about the Chinese company or government entity that shipped M-11 missiles to Pakistan and now, according to press reports, is helping Pakistan build a factory for the indigenous manufacture of M-11 missiles? Why isn't that entity on the list? What about the Russian company or government entity helping Iran to upgrade its nuclear program and ballistic missile programs, why aren't they on the list?

Madam President, this list does not solve the problem. If anything, it makes it more confused, it makes it more difficult for American exporters to determine who should or should not receive American high-performance computers. In many ways, this list is worse than nothing.

There are many who believe the entire high-performance computer export control policy of this administration is a failure. However one views this policy as a whole, there is one aspect of it that we know is not working and it can be fixed now.

We know that American supercomputers are now in Chinese and Russian nuclear weapons labs. We know that they should not be there. We know that our Government, with the resources of the intelligence community, is better able to determine the identity of end users and end uses than is industry. Industry has no way to be able to determine the end use and user of its products to the degree of confidence that our intelligence agencies can do.

Right now we have the opportunity not to impose new restrictions on our supercomputer manufacturers but to shift the burden of making end-use and end-user determinations from industry to Government.

Look at this chart again and you will see that we are talking about only a very small part of the overall policy. The entire chart describes the policy and shows the number of tiers, 1 through 4, the varying capabilities on the basis of millions of theoretical operations per second, MTOPS, along the left side. And the only part of the entire export business of American supercomputers that is affected by this amendment is this part shown in the diagonal lines. The fact is, we are talking about only 6.34 percent of supercomputer exports under this policy that will be affected by this amendment.

The Cochran-Durbin amendment will not prevent a single supercomputer export to anyone who should have one, but it will help ensure, though, that only those who should have them will have them. The only supercomputer sales that would be blocked by our amendment are those going to foreign entities who the U.S. Government determines shouldn't have it. It will not prevent legitimate sales to legitimate users in the U.S. or outside the U.S., but it will help prevent a repeat of the errors that have allowed American supercomputers to go to Russia and to go to China and be used in their nuclear weapons labs.

Let's be clear what this debate is about. It is about U.S. national security. If you think Russia and China shouldn't be using American supercomputers to improve the quality of their nuclear arsenals and the quality of the weapons systems and components and technology that they are selling in turn to others, vote for the Cochran-Durbin amendment.

President Clinton was right when he called the proliferation of weapons of mass destruction `an unusual and extraordinary threat to the national security, foreign policy, and economy of the United States,' and that it constitutes a `national emergency.' These weapons, delivery systems and technologies are more readily developed and enhanced by high-performance computers, and who makes those computers? The United States.

If the United States is going to demonstrate that it is serious about this issue, we must do more than complain to Russia and China every time one of those nations engage in proliferation.

The American fight against proliferation must start at our own borders.

I urge Senators to vote against the Grams-Boxer substitute and support the Cochran-Durbin amendment.

Madam President, I ask unanimous consent that the distinguished Senator from New Hampshire [Mr. Smith] be added as a cosponsor to our amendment.

The PRESIDING OFFICER (Ms. Collins). Without objection, it is so ordered.

[Page: S6897]

Mr. DURBIN addressed the Chair.

The PRESIDING OFFICER. The Senator from Illinois.

Mr. DURBIN. I rise to speak on behalf of the same amendment which my colleague, the Senator from Mississippi, Senator Cochran, has just described.

I am happy to join him as a cosponsor on this important amendment. I only wish my colleagues and many others who are listening to this debate could have been there when Senator Cochran's subcommittee met just several weeks ago and really talked in depth about what we are doing.

For the average layman, the average person in the United States, there are some very technical terms involved in this debate. But the purpose of this amendment is very clear and very straightforward. We understand that if we give to another country certain information or technology, they are able in many ways to use it for positive reasons. We fear however that if that same information and technology is given to a country which might use it for negative purposes, that it is inconsistent with the national security of the United States.

The Cochran-Durbin amendment is an effort to make certain that we continue to sell technology around the world, but take care not to sell it in those countries where it may be misused.

Unfortunately, the Clinton administration over the last years has had a change in its policy, with a more expansive, more liberal trade policy when it comes to supercomputers. It has been my fear, and the fear of the Senator from Mississippi, that some of these computers which are being purchased for nominally peaceful reasons are in fact going to be used for military purposes.

One of the examples which the Senator from Mississippi used in closing was the whole question of weapons testing. Some 35 years ago when President Kennedy spoke to the Nation, he challenged us as a world to reduce nuclear arms testing so as to make this a more peaceful planet. I think President Kennedy was right. And I support a weapons test ban. I think the United States should continue to show leadership.

But we live in a different world some 3 decades later where a country with a new computer, the supercomputers that we are describing, that country may have the capability to test a nuclear weapon without ever detonating it. They can set up all of the parameters within the computer, test the weapon, and show its impact.

So if you are talking about reducing the proliferation of dangerous weapons--nuclear, chemical, and biological weapons--you must necessarily get involved in this debate, which Senator Cochran has initiated and I have been more than happy to assist in.

Some questions have been raised. And I wonder, just for purposes of clarification, if I could ask Senator Cochran a question or two for the record here. I know the Senator has covered most of this in his opening statement, but I think we ought to make a clear record for our colleagues on the amendment.

One of the first things that is said is, well, you set the standard too low. If a company wants to sell this computer, which we describe as a 2,000 MTOPS computer, you have set it too low, set it at a standard so that the computers that are going to be licensed, there is going to be surveillance at such a level. It will not hit the ordinary business computers.

I would like you to respond. And I know you did respond in the course of your opening remarks to that particular criticism. If you would, please, I yield to the Senator.

Mr. COCHRAN. If the distinguished Senator would yield, I appreciate very much not only his question but also his very helpful involvement in this issue and cosponsoring the amendment.

But he gets to the central point of the debate here. It is not that this amendment sets any new levels of prohibition or granting authority for export sales. It does not change any of those levels. The level that is established by the administration is the 2,000 MTOPS level. We do not change that for tier 3 countries, as demonstrated in the chart I showed a while ago.

We were told in our hearing that 250 MTOPS is about the current power of a PC which is sold in the market here in the country now. And that under the so-called Moore's Law that doubles every 18 months. So it would be 4 1/2 years before you get to a level where you would even reach the 2,000 MTOPS level which is the trigger level for tier 3 countries that have to have a license if the end use or the end user is military. If they're civilian, you do not have to have a license at all.

What this amendment changes is who determines the end use or the end user. Our amendment says it should be the administration's responsibility. Current policy is that exporters have the responsibility of making that determination. That is the only thing we change.

Mr. DURBIN. If I could pose another question to my cosponsor on this amendment, Senator Cochran.

There have been others that have said, well, why is the United States doing this? If we stop selling computers around the world, whatever their capability, some other country is going to sell them. So we are tying the hands of American business in a futile effort to stop this march of technology.

I would appreciate it if my colleague, the Senator from Mississippi, would address that particular complaint.

Mr. COCHRAN. Our information, derived at our hearings through expert witnesses, was that we have the highest capability of any country in the world in terms of supercomputer manufacturing technology. We manufacture the state-of-the-art supercomputers. We do not have any competitors. Japan manufactures some high-performance supercomputers but their export policy is more restrictive than ours. They require licensing, we do not.

What we are suggesting here is that the policy of our administration is flawed in that it ought to make the determination in those questions where end use and end user is relevant as to whether you can or cannot make the sale, the Government ought to monitor and verify that this sale is permissible. And it applies to only 6.34 percent of the total computer sales of all American exporters in the export market.

Mr. DURBIN. I thank my colleague.

I think he noted in the course of his remarks that last week or perhaps the week before the administration said, well, let us put out a list of 13 or 14 different entities that we think we should take care not to sell to. And I agree completely with the Senator from Mississippi that it is hardly a comfort in this argument that we are protecting the interest of the United States with this list.

It is hard to believe that our intelligence operations would make a complete disclosure of every potentially bad purchaser around the world without in fact disclosing very sensitive classified information. It is far better to take the approach which the Cochran-Durbin amendment does, which says that on a case-by-case basis there will be a license issued by the Government to determine whether the would-be purchaser in any way raises a suspicion that this technology is going to be misused, used against the United States.

I think our approach to it gives the Government the power it needs to police the sales, says to the seller, the computer company, you can come to the Government now and entrust that decision to an entity which should know as to which purchasers should not be trusted. And that I think would give the industry some peace of mind. It has to be a major embarrassment to these companies to realize now that they have sold these supercomputers in China and in Russia and that they may be used for military purposes against the United States.

Certainly, these companies in the United States value our security, they are as patriotic as many others, and they would want to do the right thing. The Cochran-Durbin amendment sets up I think a good framework for the right decision to be made. I certainly hope that when this amendment comes up for consideration that many of our colleagues on both sides of the aisle will stop and pause and reflect on it. Because I think it in a way takes a look at the world as it currently exists and says we do not want to sell to potential enemies or to suspect nations that power that might come back someday to haunt us. It is important to increase trade, but not at the expense of the security of the United States.

I thank my colleague from Mississippi for his leadership. And I am happy to join him in this effort.

I yield back.

[Page: S6898]

Mr. COCHRAN. Madam President, I ask unanimous consent that there be printed in the Record a chart on exports of high-performance computers; and an unclassified report from the Director of Central Intelligence, as mentioned in my earlier remarks; and an editorial from the St. Louis Post-Dispatch suggesting that the administration should not wait, that it must act now on this issue.

There being no objection, the material was ordered to be printed in the Record, as follows:


Number of systems by country




















Czech Rep.














Hong Kong




















Korea, South










New Zealand


















Saudi Arabia




Slovak Rep.




S. Africa
























Total number of systems


The Acquisition of Technology Relating to Weapons of Mass Destruction and Advanced Conventional Munitions


The DCI submitted this biannual report in response to a Congressionally directed action in Section 721 of the FY 1997 Intelligence Authorization Act:

`(a) Not later than 6 months after the date of the enactment of this Act, and every 6 months thereafter, the Director of Central Intelligence shall submit to Congress a report on

(1) the acquisition by foreign countries during the preceding 6 months of dual-use and other technology useful for the development or production of weapons of mass destruction (including nuclear weapons, chemical weapons, and biological weapons) and advanced conventional munitions; and

(2) trends in the acquisition of such technology by such countries.'

At the DCI's request, the Nonproliferation Center (NPC) drafted this report and coordinated it throughout the Intelligence Community. As directed by Section 721, subsection (b) of the Act, it is unclassified.


The threat from the proliferation of weapons of mass destruction and missiles is one of the highest priorities for intelligence. In the US effort to counter weapons proliferation, the Intelligence Community has taken an active role in supporting US government initiatives to strengthen export controls in supplier countries and to work with other countries to prevent the sale of weapons of mass destruction (WMD), advanced conventional weapons, and their related technologies. While it is an extremely difficult problem, US government efforts have made some progress, making both the acquisition and development of WMD more difficult and costly for proliferators.

Interdiction of WMD and the technologies necessary to acquire a WMD capability is a key component in the acquisition prevention effort. We see interdiction efforts falling into three basic categories:

Preventing the transfer of materials through export controls and international nonproliferation regimes;

Halting the transfer or the negotiation of transfer of materials through diplomatic and liaison initiatives;

Seizing proscribed materials in transit, through law enforcement agencies in cooperation with the Intelligence Community.

Interdiction efforts are an extremely important part of our overall nonproliferation strategy. By themselves, however, they generally do not get countries out of the business of proliferation. They do, though, buy time for other initiatives that may be more successful in halting or rolling back a WMD program. These other initiatives can include:

Diplomatic efforts designed to reduce the perceived need for a WMD capability;

Education efforts to show that WMD-related funds would be better spent elsewhere;

Bilateral or multilateral incentives. Such incentives could be financial, including membership in an international economic forum, in exchange for halting or rolling back a WMD program;

Military assistance or security guarantees.

The US clearly leads the way in programs in all three classes of interdiction efforts. US export license applications of concern are scrutinized by a number of agencies, including the Intelligence Community. The US also is developing procedures to share appropriate end user information with key allies in an effort to strengthen our mutual export control activities. In addition, the
procedures for alerting other governments of impending transfers and tracking resulting actions are in place and working. Interdictions of shipments are occurring.

An example of a successful interdiction would be the seizure of chemical precursors destined for Libya. Although such a seizure would not halt Tripoli's aggressive chemical weapons development program, at a minimum it would:

Slow Tripoli's ability to begin serial production of chemical agents;

Provide the US time to persuade supplier nations or companies to halt future shipments to Libya;

Allow the Intelligence Community and US law enforcement agencies to identify and target new intelligence sources that could contribute to rolling back Libya's CW program;

Increase the cost to Libya of its CW development program.

Interdiction successes rest, in large measure, not on the quantity of information available to the policymaker, but on the quality. This is true for all three classes of interdictions. In licensing, for example, policymakers need unambiguous intelligence information before making a decision to deny a license, thereby denying a sale for the US company. Likewise, demarches to other governments must be accurate or the US will be accused of crying wolf and lose support from even friendly countries. And interdictions of shipments in transit often become international incidents, and potential embarrassment if the targeted material is not found in the shipment.

Actionable intelligence in support of interdiction efforts requires more than cooperation between US intelligence, policy, and law enforcement agencies. It demands close working relationships between the United States and other foreign governments committed to halting the proliferation of WMD. Such relationships will, of course, include intelligence sharing arrangements, but equally important are diplomatic, military, and scientific exchanges at all levels.

As noted above, interdiction programs by themselves cannot halt the proliferation of WMD. Alternative suppliers and technologies, increasing use of denial and deception, and a growing ability to produce indigenously weapons or their component parts are opening new avenues to states or organizations determined to obtain a WMD capability. The increasing diffusion of modern technology through the growth of the world market is making it harder to detect illicit diversions of materials and technologies relevant to a weapons program.

We are addressing these new challenges with more aggressive efforts, which go beyond traditional cold-war efforts aimed merely at understanding weapons and associated plans. We are better integrating technical analysis with political, military, and diplomatic analysis to provide policymakers with information on the motivations that drive foreign actions and decisions, and on
influential opposition forces that could support initiatives to diminish or eliminate the proliferation threat.

Our concerns are not limited to interdicting materials and technologies to state-sponsored WMD development programs. As worrisome, in our judgment, are terrorist groups and cults that seek to acquire or develop chemical and biological weapons on their own. For example, the incidents staged in March 1995 by the Japanese cult Aurn Shinrikyo demonstrate the use of WMD is not longer restricted to the battlefield. Terrorist groups and violent sub-national groups need not acquire a massive infrastructure to create a deadly, arsenal. Only small quantities of precursors, available on the open market, are needed.

Interdiction efforts are further complicated by the fact that most WMD programs are based on dual-use technologies and materials that have legitimate civilian or military applications unrelated to WMD. For example, chemicals used to make nerve agents are also used to make plastics and to process foodstuffs; trade in those technologies cannot be banned.

Nonproliferation regimes provide international standards to gauge and address behavior. They provide diplomatic tools to isolate and punish violators. The past few years, many states have joined these regimes and outsiders are encountering new pressures to join. Procurement costs have risen because of the need for convoluted efforts to hide purchases. That said, these regimes can be deceived by determined proliferators. The sheer volume of international commerce, increased self-sufficiency, and the global diffusion of technology and its dual-use nature make the regimes' road ahead a difficult one. Intelligence will play an increasingly important role in maintaining their effectiveness. Protecting sources throughout this process will be a challenge.

Following are summaries by country of ACW- and WMD-related acquisition activities (solicitations, negotiations, contracts, and deliveries) that occurred between 1 July and 31 December 1996.

[Page: S6899]

Acquisition by Country

We chose to exclude countries that already have substantial ACW and WMD programs such as China and Russia, as well as countries of lower priority that demonstrated little acquisition activity of concern.


During the last half of 1996, Egypt obtained Scud-related ballistic missile equipment from North Korea and Russia.


India sought some items for its ballistic missile program during the reporting period from a variety of sources. It also sought nuclear-related items, some of which may have been intended for its nuclear weapons program.


Iran continues to be one of the most active countries seeking to acquire all types of WMD technology and advanced conventional weapons. Its efforts in the last half of 1996 have focused on acquiring production technology that will give Iran an indigenous production capability for all types of WMD. Numerous interdiction efforts by the US government have interfered with Iranian attempts to purchase arms and WMD-related goods, but Iran's acquisition efforts remain unrelenting.

For the reporting period, China and Russia have been primary sources for missile-related goods. Iran obtained the bulk of its CW equipment from China and India. Iran sought dual-use biotech equipment from Europe and Asia, ostensibly for civilian uses. Iran was actively seeking modern tanks, SAMs, and other arms from the Commonwealth of Independent States (CIS), China, and Europe. Besides some large projects with China, Iranian nuclear-related purchases were not focused on any particular countries and were only indirectly related to nuclear weapons production.


We have not observed Iraq purchasing advanced conventional weapons or WMD-related goods, although it has purchased numerous dual-use items.


Despite the UN embargo, Libya continued to aggressively seek ballistic missile-related equipment, materials, and technology from Europe, the CIS, and the Far East. CW-related purchases diminished, however.


North Korea's WMD programs are largely indigenous. We observed no significant procurement involving ACW or WMD-related goods.


Pakistan was very aggressive in seeking our equipment, material, and technology for its nuclear weapons program, with China as its principal supplier. Pakistan also sought a wide variety of nuclear-related goods from many Western nations, including the United States. China also was a major supplier to Pakistan's ballistic missile program, providing technology and assistance. Of note, Pakistan has made strong efforts to acquire an indigenous capability in missile production technologies.


Syria continued to seek CW- and Scud-related goods during the reporting period. Russia and Eastern Europe were the primary target for CW-related purchases, while North Korea and Iran have become important suppliers of Scud-related equipment and materials.

Key Suppliers


During the last half of 1996, China was the most significant supplier of WMD-related goods and technology of foreign countries. The Chinese provided a tremendous variety of assistance to both Iran's and Pakistan's ballistic missile programs. China also was the primary source of nuclear-related equipment and technology to Pakistan, and a key supplier to Iran during this reporting period. Iran also obtained considerable CW-related assistance from China in the form of production equipment and technology.


Russia supplied a variety of ballistic missile-related goods to foreign countries during the reporting period, especially to Iran. Russia was an important source for nuclear programs in Iran and, to a lesser extent, India and Pakistan. Russia also negotiated the sale of advanced weapon systems, such as the SA-10 to Cyprus, and is an important target for Middle Eastern countries seeking to upgrade and replace their existing arms.


North Korea continued to export Scud-related equipment and materials to countries of concern during this reporting period.


Among Western nations, Germany was the favorite target for foreign WMD programs. German export controls were effective in thwarting many of these attempts, but some dual-use goods were exported, purportedly to civilian end users.


Despite our efforts, countries of concern continued last year to acquire substantial amounts of WMD-related equipment, materials, and technology, as well as modern conventional weapons. China and Russia continued to be the primary suppliers, and are key to any future efforts to stem the flow of dual-use goods and modern weapons to countries of concern.

Countries determined to maintain WMD programs over the long term have been placing significant emphasis on securing their programs against interdiction and disruption. In response to broader, more effective export controls, these countries have been tying to reduce their dependence on imports by developing an indigenous production capability. Many Third World countries--with Iran being the most prominent example--are responding to Western counterproliferation efforts by relying more on legitimate commercial firms as procurement fronts and by developing more convoluted procurement networks. Should countries such as Iran ever become self-sufficient producers and exporters of WMD-related goods and conventional weapons, however, opportunities to prevent acquisition will be dramatically limited.



From the St. Louis Post-Dispatch, July 6, 1997


China's Dangerous Computer Diversion

The Chinese have done it again--diverted machinery supposedly purchased for commercial purposes to military uses. Predictably, China denies all, but the U.S. State and Commerce departments say they have proof that China diverted a supercomputer that can be used to upgrade military hardware. The Clinton administration is rightly calling attention to the problem, but may have been lax in allowing it to happen in the first place.

Supercomputers can process so much data so quickly that any nation possessing one can significantly upgrade its weapons. That's why sales of supercomputers for military purposes require a license. But under a Clinton edict adopted in 1995, sales of supercomputers for commercial purposes don't. That appears to have been a mistake.

U.S. officials have discovered that a supercomputer manufactured by Sun Microsystems was sold to a Hong Kong company, then purchased by the Chinese government. It was supposed to be sent to a science institute in Beijing, but ended up instead in Changsha where it is being used for military applications, the U.S. says.

China denies it, as it also rejects State Department charges that it has been selling nuclear and ballistic missile technology to Pakistan and Iran. These wouldn't be China's first untruths; last year, China diverted a huge metal stamping machine sold by McDonnell Douglas for commercial airline manufacture to military use.

All supercomputers are capable of so-called dual use, that is, of being employed for both peaceful and military purposes, so they must be carefully monitored. Though the United States has been fairly successful in that effort with its sales to Russia, China has been largely uncooperative. Congress is so concerned that the House has passed a bill reinstating the requirement that all supercomputers sold abroad for any purpose be licensed--and their use be tracked.

In 1995, the administration deregulated the sale of supercomputers for peaceful purposes on the ground that if America doesn't sell its machines, the Europeans or the Japanese would sell theirs. But the importance of slowing the spread of higher grade nuclear weapons and ballistic missiles requires the U.S. to prevent the sale of supercomputers which defeat that purpose, never mind helping the computer industry compete abroad. Only strict licensing is safe, and our competitors should be pressured to follow that policy. The administration shouldn't wait for Congress, but require it now.

Mr. LEVIN addressed the Chair.

The PRESIDING OFFICER. The Senator from Michigan.

[Page: S6900]