Dual-Use Technologies and Export Control in the Post-Cold War Era (1994)

Dr. Seymour E. Goodman

Department of Management Information Systems

University of Arizona

The fifth point of the Protocol proposing bilateral activities on the problems of dual-use technologies and conversion calls for in-depth discussions of "case studies of programs to develop and to apply dual-use technologies for civilian purposes which are also designed to limit diversion of such technologies to military applications." [1]

Since the development and analysis of case studies requires substantial time and effort, it is necessary that we start our discussions by seeking agreement on the reasons for pursuing such studies and on how we would expect to benefit from them. From our meetings in Russia in December 1992, it would seem that both sides at least tentatively agree

• that some dual-use technology and product transfer between the two countries is necessary and desirable for important economic and scientific reasons;

• that it is desirable to reconsider the Soviet-era assessments of the risks of trade in dual-use technologies with Russia and the other former republics of the former USSR, and that trade controls be adjusted accordingly if the risk assessments are revised;

• that it is necessary for both countries to work together to limit diversions of dual-use technologies for military applications by third countries.

Progress along these lines arguably will depend on both sides cooperating to

• reduce many of the uncertainties in American export control policies and practices so that the United States will become a more predictable and reliable partner in the trade and transfer of dual-use products and technologies;

• help build a new legal, institutional, and behavioral order in Russia that will be able to insure against diversions to the same degree that is acceptable for most other countries;

• develop a new set of relations between the United States and Russia that will enable them to work together to prevent and track undesirable diversions of dual-use products and technologies.

To these ends, case studies would hopefully serve at least three important complementary purposes:

1. as test beds, for ascertaining the feasibility of, and to work on solving the practical problems that will arise with, control mechanisms such as verification schemes;

2. to develop models, consolidating the knowledge and experience obtained from simple cases and other experience into prototypes or models for effective vehicles to promote the use of dual-use technologies for civil purposes while controlling military diversion (these models might eventually be explicitly packaged into licensing mechanisms and conditions.);

3. for confidence building, using a small number of relatively controlled and risk minimizing cases to help build experiences that will be necessary to increase the confidence of both sides in each other and in their prospects for working out longer term and more extensive relations.

These three purposes apply to Russian-American trade and technology transfer and to trade and technology transfer between either and third countries. Clearly, case studies also serve as vehicles for the analysis of problems and the implementation of solutions that arise under many of the other points listed in the December 1991 Protocol; in particular, to the points relating to confidence building (point 3 on openness in research, point 6. on industry reporting, and point 9. on control and verification measures). [l]

If we agree that case studies would prove valuable, then the next step is to decide what to study. Some possibilities that immediately suggest themselves are case studies of one or more of the most important dual-use technology sectors, such as computers, advanced materials, or civil aircraft. A key question that might be examined sectorially or more generally is the extent to which the former Soviet defense industries can effectively convert or spin off civil enterprises that absorb imported dual-use technologies and yet insure against diversion to sister enterprises still in the weapons business. One can study the forms of past diversions, the effectiveness of organizations such as joint ventures to help prevent diversion and provide verification, technological "choke points" [1], and the technical and economic factors affecting the controllability of products. Some of this has been done before [2, 3, 4], and the prospects of cooperative Russian

American efforts greatly increase the expectation of more complete studies that might be constructively used as discussed above.

We should also consider a somewhat more general case study of the changing institutions and priorities in Russia. Much of the practice of U.S. export controls against the USSR was based on a model of the Soviet system in which great priority and authority were given to Soviet security agencies. Under this model, it was assumed that the KGB and military-industrial organizations could obtain access to any foreign technology or products on Soviet territory, and that Soviet civil enterprises were almost powerless to restrict this access. This model generated a lack of trust in the ability of the Soviet legal or administrative system to prevent diversion or other misuse of dual-use technologies, and was used to justify very conservative U.S. export control policies and practices. One particularly important consequence of this distrust was an export control policy that amounted to a form of de facto economic warfare against Soviet civil industry.

Today's Russia is not the same place the Soviet Union was a decade ago, and U.S. policy has changed in some fundamental ways. We now want the Russian civil economy to develop in ways that will support a more democratic and consumer oriented civil society. To this end, we recognize that large parts of the former Soviet defense industry must be converted or dismantled and replaced by modem and viable state and private civil enterprises. For this to happen a necessary, but by no means sufficient, condition is for more normalized technology transfer and trade relationships to develop between Russia and the rest of the world, and this will require modification of old export control policies. This, in turn, will require confidence on the part of the United States that the former Soviet model for priority and authority has been replaced by a legal and administrative system in Russia that can be trusted to protect civil technology acquisitions from diversion for military purposes.

Accordingly, it is necessary for us to learn more about the current and changing legal and administrative structures in Russia, and to obtain some experience with possible methods of export control in the context of this environment. We must study the extent to which civilian and military activities are tied together organizationally and financially. We have to develop a more up-to-date model of the internal situation in Russia and of how this affects questions of access to and diversion of dual-use technologies. This applies to both the import of foreign technologies and to the export of Russian or the re-export of foreign technologies to third countries. Ideally, we would eventually like to have enough confidence in the Russian civil authority to permit the United States. and the other COCOM countries to deal with Russia in the way they deal with most other countries. This is arguably the most important case study we might undertake. Such a study might be undertaken broadly, or as a small pilot exploratory study to obtain some initial experience and data from a particular industry.

Another possibility would be to consider a case study of an important verification mechanism, such as end-use controls [4, Appendix C]. These are controls to limit the

misuse of products that are physically located on foreign territory. Several forms of misuse are possible:

• inspection and analysis with the intent to duplicate or find the weaknesses of the product.

• diversion-in-place; for example, the use of a supercomputer located in a weather research facility to run military-related applications.

• relocation and diversion; for example, the movement of an array processor from a scientific laboratory to a submarine.

• diversion of manufacturing capability; for example, the use of imported microelectronics manufacturing equipment to produce components to be used in weapons systems.

End-use controls are most effective when applied to so-called "high walls" products that can be located, traced, observed, or otherwise tracked on an individual basis in a protective environment [4, Appendix C]. Some examples of such products include large supercomputers and major pieces of semiconductor manufacturing equipment. They also tend to be most effective against misuses in the form of relocation and diversion. The effectiveness of end-use controls against other forms of misuse is strongly dependent on specific local conditions.

A case study of end-use controls would consider the full spectrum of specific kinds of end-use controls and the circumstances in which each would be most effective. It would also consider different means of administering such controls.

As a final example, we might consider a case study of possible joint mechanisms and institutions for export controls. A joint Russian-American commission would be established to look into questions and procedures that could be applied to the export of dual-use technologies from either to certain third countries. The focus of this case study would be to carefully examine the current practices in both countries for licensing and monitoring the export of dual-use technologies in order to find common ground for a joint control regime. The joint commission would analyze several instances of real exports and some diversions in detail.

I would recommend that we attempt a broadly exploratory case study that would at least start to look into several of the topics sketched above, but with special emphasis on confidence building in multiple ways. To keep this pilot project manageable, we

should try to obtain our initial experience and information from a fairly small, but very important, specific technological sector.

High performance computing (HPC), including both traditional and massively parallel supercomputers, would be an excellent candidate for such a case study. Given our interests in the control and conversion of dual-use technologies, this sector might be usefully extended to include large mainframes and high performance networks as well.

HPC has been an exceptionally important and visible technology sector with a long history in the export control arena. Applications span an extraordinary spectrum of scientific, industrial, and military uses. Many of the most difficult problems of the control and conversion of dual-use technologies are encapsulated in this small sector. Both the United States and the former Soviet Union have exported large computers. A good case study with a focus on HPC should attract considerable interest.

The study of HPC, and of supercomputers in particular, should be manageable for at least four reasons. First, as compared with most other industries, a relatively small number of facilities are engaged in the development and manufacture of such machines. Second, almost all of these facilities are in only three places: the United States, the former Soviet Union, and Japan. Third, this is a "high walls" technology with a fairly small number of large end products that can be tracked as individual units on a worldwide basis. Finally, HPC has been the object of both general and specific controls and control procedures which might be usefully reviewed with respect to transfer to Russia, and which may serve as "strawmen" for proactively trying to help create a new civil authority for the control of dual-use technology in Russia and for exploring some specifics of joint Russian-American control regimes.

A case study of HPC could serve as a vehicle for testing arrangements, developing models, and building confidence in at least the following ways:

The number of facilities engaged in the development and production of large computers in Russia is small enough for us to do a semi-exhaustive survey of the post-USSR movement of technical specialists in this technology. We might also note that at least one promising model has been worked out to keep R&D groups in this technology together and working inside Russia for a U.S. company, i.e., the arrangement for the Babayan group at the Institute of Precise Mechanics and Computer Technology (ITMVT) in Moscow. This model and other forms of joint ventures should be studied carefully. We might also look into the movement and brain drain of technical specialists in the C³ (command, control and communications) area who worked on military systems using high end computers.

As stated in Glenn Schweitzer's paper, "A Conceptual Approach to Addressing Dual Use Technologies: A Framework for US-Russian Dialogue," "If enterprises are to change from military to civilian production, they must replace their previous dependence on defense contracts with new approaches to financing, marketing, and profitability. Unless such a transition is successful and clear to all, international suspicions will linger that conversion has not really taken place and that technologies continue to be used for [military-related purposes].... [There must be] persuasive economic [and technological] evidence that conversion activities are genuine and permanent." [6] Clearly, detailed studies of enterprises in the process of conversion are necessary. HPC development and production facilities in the former USSR have had a long history of functioning under the purview of various combinations of both the Academy and the military-industrial ministries. A prime example is ITMVT in Moscow, which is under the Academy, and has had a long history of developing large machines for manufacture under the Ministry of the Radio Industry, and which has dealt with hundreds of space, military, and military-industrial customers since its creation in the early 1950s (late 1940s if one counts its predecessor in Kiev). As such, development and production facilities for high end computers should be ideal for the purposes of our case study.

From the U.S. perspective, the best kind of case study would be one in which the Russian side is willing to find out about and reveal how technology was diverted as a means to help build confidence on the U.S. side, and to aid in help devise legislation, procedures, and institutions that will help build a civil authority capable of preventing, exposing and punishing such diversions in the future. Diversions in the high end computing arena should be fairly easy to identify since they are likely to constitute a small number of very discrete cases. The Soviet-era structure of the VPK (Military-Industrial Commission) and the ministries under its purview (particularly Minradioprom, Minelektronprom, and Minpromsvyazi in the case of HPC) had much to do with the reality and perception of the likelihood and authority for diversion. Our case study of HPC would provide a useful vehicle for assessing the current state of that structure, the participation of the intelligence agencies, and the extent to which Russian enterprises have become less vulnerable to pressures to cooperate in various forms of diversion.

Supercomputers are "high walls" products in that they are usually heavy and large, produced in small numbers, maintain internal audit trails of users, and require prolonged vendor support. As such, HPC systems are well suited for consideration for verification and monitoring mechanisms such as end-use controls. A particular form of end-use controls, the Supercomputer Safeguard Plan, is one of the few examples with some serious history of use. One might argue that if we cannot work out effective verification

schemes for these products, then we are not likely to be able to work out schemes for many dual-use technologies of great consequence.

The "high walls" characteristics of large computer systems should make such products ideal for a pilot case study of how the two sides can work together on the export or re-export of dual-use technologies to third countries. We can start with a review of how both sides have controlled high end computers in the past, with special attention to how specific exports have been tracked abroad. Specific forms of cooperation for the future would be considered. It is worth noting that while most of the former Soviet HPC sector is located in Russia, some significant portions are in Armenia, Belarus, Estonia, and Ukraine, adding an important element to this aspect of our prospective case study.

These possibilities are by no means exhaustive. For example, the tremendous applicability of HPC to a broad spectrum of scientific and military problems could argue for a good case study of the problems of the separation of research for civilian and military purposes. In conclusion, the point to underscore is the prospective viability and manageability of HPC as a suitable case study for almost any purpose of interest to our joint undertaking with regard to dual-use technologies and conversion. In this vein, we also have the option for a flexible and extensible case study that might be started in a modest and limited way, and expanded as experience and funding are acquired.