Mr. INGRAM. Regardless of the legislative history, the practice that takes place now is that for matters or information pertaining primarily to military utilization of nuclear weapons, you are following the traditional executive orders approach to security classification.

Mr. SEWELL. That is correct with regard to security protection. Mr. INGRAM. So that what we are talking about then is a carveout for the Department of Defense from the 'born classified 'procedures that are followed by the Department of Energy.

Mr. SEWELL. Yes; it is really the restricted data procedures that are prescribed by the law that we needed to change to have practical procedures for the military to follow.

Mr. SILVERSTROM. In essence, if I may, it is no carve-out from the definition of restricted data. It is simply examining a given piece of restricted data to assess jointly with the Department of Defense whether that information can be sufficiently safeguarded domestically as NSI.

If so, it is given the special category FRD and safeguarded as NSI, but it remains a distinct classification category under the Atomic Energy Act. It is not a carve-out.

Mr. INGRAM. Perhaps you can elaborate on that more in your followup to us.

Mr. SEWELL. Yes.

Mr. INGRAM. The issue that I am trying to get at, though, what we really seem to be coming to as a practical matter, is how best to protect privately generated data that as a practical matter were only according security procedures to documents that are marked, which is not unlike the military executive orders NSI approach.

From a legislative or a policy standpoint, what we are grappling with here though is how to maintain some sort of control over information that is generated in the private area, not by the Government, not by the Federal contractors.

The Atomic Energy Act does restrict data generated totally outside Government funding. Has there ever been any court test as to the applicability of these provisions for privately restricted data, outside of the Progressive magazine case?

Mr. SEWELL. I will have to turn to Mr. Silverstrom. I am not aware of any myself.

Mr. SILVERSTROM. I am sorry, I am not sure I quite follow. Mr. INGRAM. Has there ever been court challenge to the Atomic Energy Act's restricted data provisions in terms of their coverage of privately generated data? I am not aware of any.

Mr. SILVERSTROM. To the best of my recollection-I believe you and I have discussed this before-the answer is no.

Mr. INGRAM. From documents provided the subcommittee, the Justice Department has on at least two occasions in memos to the old Atomic Energy Commission expressed concern about the applicability or the abilities to successfully withstand a court challenge to the act's applicability to privately generated data.

Is this a matter of continuing concern to the Department of Energy?

Mr. SEWELL. Again, let me turn to Mr. Silverstrom.

Mr. SILVERSTROM. I am not sure that there were concerns so much about the definition of restricted data or the defensibility of

the concept of restricted data, but rather some specific actions and regulations that the Atomic Energy Commission was proposing to take that perhaps extended that concept into a prohibition going beyond the very explicit provisions of the act-a view that one should take the most conservative approach and look to an act of Congress to provide a fundamental basis for an actual prohibition, rather than extending the current definition.

Mr. INGRAM. The Atomic Energy Act forecloses from patentability only inventions useful solely in atomic weapons or inventions that have multiple use as to the extent that such invention would be useful in atomic weapons.

This is a smaller circle of inventions than the broader circle of inventions falling within the definition of restricted data.

As a practical matter, is the smaller set dealing with weaponry the daily rule of thumb that the classification office applies to areas of private invention that would come within the ambit of restricted data?

Mr. SEWELL. Let me turn to Mr. Duff to elaborate on that. Mr. DUFF. I would say no because again the measure that is applied, if we are given a patent to review, is the definition of restricted data, so it is not limited to the circle of nuclear weapons.

It does include the production of special nuclear material, and certain military nuclear reactor information as described in the act in the definition of restricted data.

Mr. INGRAM. What I am leading toward is whether or not there is a narrow subset of inventions that as a practical matter are accorded secrecy protection, which is much narrower than the present definition of restricted data.

Mr DUFF. Not to my knowledge.

Mr. SEWELL. No.

Mr. SILVERSTROM. No.

Mr. INGRAM. Thank you, Mr. Chairman.

Mr. PREYER. Mr. Morr?

Mr. MORR. Mr. Chairman, I have no questions.

Mr. PREYER. I just have one other question more in the nature of a comment, I guess, Mr. Sewell.

The Progressive magazine case will have very profound implications for future policymakers faced with the publication of details of some advanced military technology. Since this is going to be a continuing and important problem, it would seem to me worthwhile to have as complete a record as possible of the various decisions that have been made in this case to pursue the litigation, the risk involved, the judgment calls, that could be an important historical study.

I wanted to ask if the department would be willing to commit itself and its history office to preparing such a case study, and begin now to collect all of the relevant internal documents from the various agency players involved.

We are not asking you, of course, to publish it, but I think this could be an important question in the future, a study of it and particularly a collection of the important documents before they slip away. It could be very helpful in the future.

Mr. SEWELL. Mr. Chairman, I think that is a very good comment. We have discussed similar thoughts and ideas among ourselves

because I agree with you that it is very much a cornerstone case, really being the first major test of the Atomic Energy Act, and therefore should be documented to the fullest to the best of our ability.

Certainly, I for one feel that is an important thing to do and will do everything within my power within the Department to see to it that we do put together such a file or compilation of items that can be used for future evaluation when questions and problems come up.

Mr. PREYER. Thank you. I think that would be very helpful. We appreciate your being with us today, Mr. Sewell, Mr. Duff, Mr. Silverstrom.

If there are no further questions, the subcommittee will stand in

recess.

[Whereupon, at 12:45 p.m., the subcommittee adjourned, to reconvene subject to the call of the Chair.]

The Atomic Energy Act contains a provision, 42 U.S.C. $2274, oriented toward classification-style government control 1/ of technical information generated outside the government. The statute goes beyond the usual classified information concept in that it may apply to information generated by an individual himself or herself through scientific investigation or inference from information in the public domain. A similar statute has 2/ been suggested for the cryptographic area. In contrast to the traditional central (but not necessarily exclusive) concern of the espionage laws - the control of information generated within the defense community, usually about specific defense efforts this special type of provision seeks to control general principles of science or technology, replicable and possibly generated outside the defense community. This statement addresses the desirability and possible revision of such provisions."

3/

-2

Two types of argument support such provisions. The first is the national security argument. Disclosure of nuclear weapons design information or of cryptographic information can be of great harm to the U.S., no matter how that information originates. Disclosure of at least the former type of information can also be of grievous harm to society itself, as by making it easier for terrorists to construct nuclear weapons.

A second argument has been less frequently made but is perhaps of greater importance: it may prove crucial for society to avoid ever developing certain technologies, e.g., simpler nuclear weapons construction methods or conceivable genetic manipulation technologies. The risks of some technologies and the limited possibilities of control once the technologies are developed may make not developing the technologies the only practical way to control them.

These concerns clearly make it generally unwise for any publisher to disseminate information such as that which might be useful in nuclear weapons design. However, there are even greater counter concerns, making government restrictions on the publisher also unwise. The first is the First Amendment tradition, a tradition which weighs especially heavily against prior restraints on the press, New York Times Co. v. United States, 403 U.S. 713 (1971); Near v. Minnesota ex rel Olson, 283 U.S. 697 (1931). While it is conceivable that the disclosure of "born-classified" information could be discouraged through criminal penalties rather than through prior restraint, it seems likely that prior restraint would be more effective, but this approach poses problems as well. Under the prior restraint approach, as applied in United States v. The Progressive, Inc. 467 F. Supp. 991 (W.D. Wis., 1979), the chances are good that after an