April 12, 2000

Mr. Chairman and distinguished members of the subcommittee, good morning/afternoon.

I'd like to make a few opening remarks, after which I look forward to answering your questions.

As Mr. Keeney noted, since joining ISS in 1991, I have worked on some of the major espionage cases of the 1990s--the Lalas case; the Ames case; the Lipka case; the Squillacote case. All of these cases resulted in prison sentences ranging from 14 years to life.

All of these were prosecutions under 18 U.S.C. Section 794. Of these, I am most proud of the Lipka case, where I helped build an historical case where the investigation did not even begin until roughly 25 years after the crime. I also take pride in the 1998 Squillacote/Stand case, where I was part of the trial team for a two-week jury trial against a well-financed defense, which resulted in guilty verdicts on all counts and sentences of 22 years and 18 years. In connection with that trial, I was awarded last year the Attorney General's Award for Excellence in Furthering the Interests of U.S. National Security. Finally, I am, of course, proud of the Ames case, for which John Dion and I received an award from the U.S. attorney for the Eastern District of Virginia.

At the time of Peter Lee's admissions in October 1997, I fully expected that they would lead to another case in my string of Section 794 cases. But almost from the outset I encountered significant obstacles.

Within about two or three days after Lee made his admissions in early October 1997, I flew out to Los Angeles and met with prosecutors from the USAO and FBI special agents from the LA Division to discuss the case. Our office had first been briefed on the case in August 1997, when it was still just a false-statements case because Lee had merely admitted to telling lies. In my trip in October, I spent several hours meeting with then-AUSA Jonathan Shapiro, and FBI special agents Gil Cordova and Serena Alston, at the LA Division FBI office, where we also listened closely to the tapes of the October interviews. To the best of my recollection, it was then that I first learned that the information Lee had compromised in 1985, while classified "Secret" then, was no longer classified in 1997, and that the information Lee compromised in 1997, was, for the most part, only classified under a mosaic theory and only at the "Confidential" level. By mosaic theory, I mean that the items of information considered separately are unclassified, but when grouped together they become classified.I also recall that, with respect to the 1997 compromise, the FBI in Los Angeles showed me a copy of a 1995 document authored by Lee that was marked "Confidential." It concerned research into detecting the wakes of surface ships, conducted under DOD auspices, through the use of radar directed at the ocean surface. Although the overall document was classified "Confidential," every single portion of the document was separately marked "Unclassified," with one exception. The exception was the single paragraph on the first page that explained that considered as a whole the document was "sensitive."

Later, after I returned to Washington, I obtained tapes of Lee's October confession and determined that as to the 1997 compromise, the 1995 "Confidential" document essentially contained all the significant information Lee had confessed to giving the Chinese in May 1997, with one important exception. The 1995 document was all about using radar to detect surface ship wakes; it said nothing about using radar to detecting submarines or anything below the surface. I knew that Lee had admitted to the FBI that he told the Chinese in May 1997 that the radar technique discussed in the 1995 document could be used to detect submarines, although he minimized the disclosure by telling the FBI that the Chinese already knew this.

In my estimation, both then and now, the sole weakness in the case was the questionable significance of the information Lee compromised, both in 1985 and in 1997. As to Lee's 1985 disclosure, I knew, for instance, that the Department had never prosecuted a case under 794 where the compromised information, as in the case of Lee's 1985 disclosure, had been declassified prior to the crime being discovered. Let me emphasize this: the information Lee admitted disclosing in 1985 had been declassified. While some aspects of the government's research in this area might remain classified, as shown by updated classification guides, what Lee confessed to disclosing regarding ICF research in 1985 was fully declassified by 1993. And on this issue, I would refer the subcommittee to the FBI's October 15, 1997 interview of Dr. Roy R. Johnson, of Lawrence Livermore National Laboratory.

Furthermore, what I later determined was that the information was actually declassified over the 1990-93 time period, not just in 1993. DOE documents that I believe this subcommittee has show that ICF research, including details disclosed by Lee to the PRC, began being declassified on March 21, 1990, for reasons that included the fact that the rest of the world was catching up. Another reason for the declassification, I was told, was that DOE considered it tobe in the U.S. national interest to educate countries on how to simulate nuclear weapon explosions in a laboratory setting, in order to discourage them from actually detonating nuclear devices. Moreover, I was advised, and again this is documented, that the debate over declassification had begun at least as early as January 1989, only four years after Lee's disclosures.

Why is any of this relevant? Why does it matter that the information was declassified after the crime? Because section 794 does not penalize disclosures of classified information. It does not even use that term. What it penalizes is the disclosure, or attempted disclosure, of items, documents and information related to the national defense. And what the caselaw, including Supreme Court caselaw says is that this is a jury issue, not to be decided by a classifier merely testifying that certain information is or was classified at the time of the offense. The government needs to be able to describe how a disclosure of classified information might benefit an enemy of the United States. And publicly available information that tends to suggest that the classified information is not all that significant may well be found by a court to be relevant and admissible in an espionage prosecution.

The DOE documents indicated to me that there would be a significant issue at any trial whether the ICF disclosures Lee made in 1985 related to the national defense at the time he made them. Most alarming to me was the notion that Lee could claim that he made the disclosures to encourage China not to conduct nuclear weapons tests in the field, and he would likely be supported by internal government documents or even testimony of former USG or Livermore officials that that was actually one of the reasons the U.S. government declassified the information beginning in 1990. In other words, Lee would have been able to argue his actions were in the national interest.

I soon discovered that there were similar obstacles to bringing a Section 794 prosecution based on the 1997 disclosure. To analyze this, it is helpful to begin with the 1995 "Confidential" document, every last substantive part of which, when considered independently, is unclassified.

Recall that this document discusses a radar technique in which the wakes of surface ships can be detected by bouncing radar signals of the ocean surface. I have a copy of it right here today.

The best way to explain the problem with basing a prosecution on this document is as follows. Under the classification guidance on this document, I could remove any single paragraph, perhaps even a single line--just cut it out--and then take the remainder of the document over to that press table, and I would not be guilty of any crime. I would not even be committing a security violation, because the document is only classified when considered as a whole.

I recognized that problem with the 1997 compromise as soon as I got to Los Angeles.

But there was one crucial piece of Lee's admissions that I thought, at the time, could make the case viable, even viable under section 794. Lee had confessed to telling the Chinese scientists that the technique described in the document could also be used to detect submarines. As I've said, that goes beyond the document. Surely, I thought, it must be a well-kept secret that the U.S. government is investigating the detection of submerged submarines by utilizing radar aimed at the ocean surface.

When I returned to Washington, as I said, I began analyzing the confession in some detail. Approximately two weeks after returning, on October 23, 1997, I attended a meeting at the Main DOJ building with the FBI and other Criminal Division attorneys, along with Mr. Shapiro and his supervisor, then-FAUSA Richard Drooyan. The problems with the information, which I've just described, were discussed, along with other issues in the case. Immediately after that meeting, I attended a briefing by the FBI on the case, along with Mr. Shapiro and I believe Mr. Drooyan. I will not go into that briefing here in open session.

A few days after that meeting, I attended a meeting with DOD officials to discuss the 1997 information. I've recently been reminded, by the testimony of DOD and Navy officials to this subcommittee last month, that that meeting occurred on October 28, 1997. The main purpose of that meeting, from my perspective, was to inquire of DOD as to what publicly available information could potentially undermine an espionage prosecution for the 1997 compromise. Another issue for me was what could the government say about the program generally, in a public forum, if the case were to go to trial.

About a week after the meeting, I received a stack of public articles from DOD related to radar ocean imaging generally. One thing they also sent me was extremely surprising. Among the articles was a print, out from a Lawrence Livermore National Laboratory web site, last updated in March 1995, well in advance of Lee's 1997 trip to China. I have a copy of the print-out here. I quickly confirmed, after receiving it, that the web site was a public one and available to anyone in the world with a computer and a modem. I offer it into the record now and I'd like to read some portions of it out loud.

The title of the page is "radar ocean imaging." The first line of text states: "This project focuses on the detection by radars of surface manifestations of moving, submerged submarines."

Later, it says that as a result of "achievements" in the project, "(t)here is now no controversy within the community that radars offer any potential for this problem," that is, to detect submarines. It concludes: "(t)his program has made impressive advances in understanding and exploiting radar remote sensing of the ocean for important national defense needs."

In addition, a few days after obtaining the website printout, DOD gave me a copy of the prepared remarks of Dr. Richard E. Twogood of the Lawrence Livermore National Laboratory, presented in open session to a subcommittee of the House Armed Services Committee in April 1994. I have a copy of those remarks and I offer it into the record now, and I'd like to quote from the most significant portions: "The Joint US/UK Radar program has made important progress in the development of methods to detect submarine signatures with remote sensing radars, especially over the last two years." It also states: "We have developed new signal processing and detection techniques that, to our knowledge, have never been successfully applied to this problem. We have applied these new methods in both classified and unclassified settings. Results have been achieved that I believe are not only impressive, but also offer great promise for future improvement."

So there it was. There was no secret at all that the USG was working on a program to detect enemy submarines with radar aimed at the ocean surface. There was not even any secret that we had achieved a potential breakthrough. The website and Dr. Twogood's testimony, coupled with the fact that the underlying 1995 document was only classified under a mosaic theory, convinced me that there was no section 794 case on the 1997 compromise.

I arrived at that conclusion even before I received the Shuster memorandum of November 14, 1997. The memo only served to reaffirm my position. Particularly significant were the Navy's determination that it could not support the "Confidential" classification of the 1995 document and that, in any event, Peter Lee's disclosures did not cause significant damage.

I would note that the Shuster memorandum had the concurrence of the vice chief of naval operations, the second highest ranking Navy official.Now, just because a compromise of classified information cannot be prosecuted under section 794 does not mean that there are no other statutes with serious criminal penalties that might apply. There are other provisions of the espionage code, specifically 18 USC 793 and 798. In addition, there is the Internal Security Act, specifically 50 USC 783. Each of these carries a ten-year penalty. The problem was that none of them applied. Section 793 was out because it too used the term national defense information, just like section 794. Section 798 was out because it applies only to communications intelligence and cryptographic information. And the Internal Security Act was out because it applied only to defendants who were USG employees or employees of USG-owned corporations. That was the biggest disappointment, and I remember discussing that with Mr. Shapiro over the phone following my trip out to Los Angeles. The statute does not apply to employees of government contractors, such as TRW.

Shortly thereafter, I do not recall precisely when, I recommended to Mr. Dion that we offer Lee a plea under 18 USC 793 or section 224(b) of the Atomic Energy Act of 1954 (42 USC 2274(b)) for the 1985 compromise. Both statutes carry a maximum penalty often years, and would require Lee to waive the statute of limitations. The USAO elected to offer Lee the plea under 18 USC 793.

At some point in early December 1997, it became apparent that Lee was balking at a plea with a potential ten-year exposure for the 1985 incident. I then recommended to Mr. Dion that, although the section 794 case for that incident had problems, it was sufficiently robust that we could still ethically use it as leverage. This was communicated to the USAO by Mr. Dion, I believe, in a phone call to Mr. Drooyan. Shortly thereafter, the plea agreement was entered. Lee did in fact waive the statute and plead guilty to a violation of 18 USC 793, along with a violation of 18 USC 1001 for lying about the circumstances of his 1997 travel to China.

It goes without saying, I hope, that I was extremely disappointed that Peter Lee was not sentenced to prison. It is the only espionage prosecution that I have worked on that did not result in a jail sentence. But let me add that I am proud of my work on the case, and proud that Jonathan Shapiro and I ensured that Peter Lee would not remain free to continue to make sensitive disclosures to foreign governments.

That concludes my remarks.