NewsMakingNews asks: Should the Monday Morning Quarterbacks who classified the data, Wen Ho Lee handled, do some meditation time at the Crow Bar Hotel?

Lee Data Constraints Unclear
By Ian Hoffman © 2000
Albuquerque Journal Northern Bureau April 10, 2000 Albuquerque Journal

     SANTA FE — Most — if not all — of the U.S. nuclear-weapons data former Los Alamos scientist Wen Ho Lee is accused of illegally copying had not been reviewed and formally classified as secret.
    Prosecution evidence shows the more than 20 weapons designs and related nuclear-blast simulations that Lee allegedly copied to portable tapes were not labeled as "restricted data" at the time.

    Restricted data is the U.S. Department of Energy classification category for secrets of designing and making nuclear weapons.

    Instead, records show Lee's tapes were full of data designated as PARD, or protect as restricted data.

    "This is ... an indication of potentially lesser sensitivity," said Steve Aftergood, a classification expert at the Federation of American Scientists, a Washington organization founded by former Los Alamos weapons scientists after the Manhattan Project.

    "It raises one more small question about the prosecution of this case," Aftergood said.

    A federal grand jury indicted Lee in December on 59 counts alleging he illegally downloaded and copied files of U.S. nuclear secrets to data tapes.

    Federal prosecutors are seeking life imprisonment for Lee. They argue he knew the files he downloaded were so sensitive as to "change the global strategic balance" and yet broke security rules so deliberately that he could only have intended to steal the data.

    Lee's attorneys have outlined a strategy of showing the downloaded bomb secrets were not that secret and that Lee had legitimate reasons for copying them.

    Neither side would comment on whether the PARD designation will play a role in the court case.

    But it does raise the question: Are never-classified data protected by national- security laws just as strongly as "top secret" and restricted data?

    "An argument can be made both ways, but the PARD designation is one step further removed from the 'crown jewels' category," said Aftergood, head of the federation's Project on Government Secrecy.

A set of rules

    Unlike restricted data, PARD is not a data classification. It is a set of rules for handling data, devised by the defunct Atomic Energy Commission so scientists would not have to classify and lock up reams of printouts in the early decades of weapons computing.

    According to a Los Alamos National Laboratory definition, PARD is "not readily recognized as classified or unclassified because of the high volume of (classified computer) output and the low volume of potentially classified data."

    On its face, DOE policy said PARD information is to be handled as if it were classified as restricted data.

    However, Los Alamos assigned PARD a lower level of computer security than restricted data. And PARD printouts and tapes could be left on scientists' desks inside high-security areas, a security violation for information classified as restricted data.

    To convict Lee on the 39 charges carrying a potential life sentence, prosecutors must prove he "removed" or "acquired" documents "involving or incorporating Restricted Data" with the intent to harm the United States or help a foreign nation.

    It is not clear whether PARD is really the same as restricted data as an element of the most serious charges against him.

    Prosecutors will try to demonstrate at his trial that the downloads were extremely sensitive — the "crown jewels" of U.S. national defense. That demonstration, aimed at suggesting Lee had a criminal intent, could be tougher if the files on Lee's tapes were not formally classified and were less protected than classified information.

    On one hand, the PARD files on Lee's tapes do contain weapons secrets that technically meet the legal definition of restricted data.

    But they had never undergone a formal sensitivity review to determine what category and level of classification applied to them.

    Those reviews came only last year, after the FBI found remnants of Lee's files on unclassified computers and some of the tapes in Lee's office outside the lab's security fence.

    Los Alamos and DOE rules require PARD to be marked by the acronym. Prosecution exhibits suggest Lee relabeled the PARD files as unclassified, so the Los Alamos network would allow transfer of the files from classified to unclassified computers.

    The tapes he created were not labeled as PARD and were found outside the lab's security fence, a violation of PARD handling rules.

    The DOE tightened the PARD rules in August and set June 30, 2002, as the deadline for the elimination of PARD. DOE officials said they wanted to get rid of the PARD designation years ago but found resistance at the weapons labs. Los Alamos and its sister weapons-design lab, Lawrence Livermore, asked DOE to keep the PARD designation until 2002 so they had time to classify and copy the data to a media such as compact disks.

Reality and legalities

    So were the files truly classified when Lee downloaded them in 1993, '94 and '97?

    DOE's Ray Holmer says all data on the Los Alamos classified network are automatically viewed as secret restricted data. Weapons data often are "born secret" and during the Cold War was so abundant it was often designated PARD as a convenience.

    "PARD was a mechanism to allow us to protect data that we knew was classified but we didn't have the resources to identify which specific pieces were classified," said Holmer, director of operations for the DOE's Office of Cybersecurity and former DOE manager of classified computer security.

    In the late 1950s, weapons scientists had no desktop computers but designed nuclear bombs on large mainframe computers using punch cards. They would check their physics formulas and calculations on reams of computer printouts. The cards, code listings and printed outputs could contain weapons secrets.

    But the printouts were too voluminous to treat as secret and lock in office safes.

    "It was useless to handle all this stuff as secret because there was too much of it. So they established something called PARD," said Bob Clark, a computational physicist who worked on weapons codes at LANL until 1995.

    "It's in-between stuff," said another Los Alamos physicist. "It's handled like RD (restricted data), but you can leave it laying around in locked offices."

    In the 1960s and '70s, code scientists might fill several boxes of PARD a week. When PARD files piled up in the hall and were declared a fire hazard, scientists were asked to set their boxes out once a week, and guards wheeled them away to an incinerator, Clark said.

    "There weren't enough safes in the DOE complex to handle all that stuff," he said. "PARD was a way to circumvent some laws we thought were too restrictive, to get some work done. At the same time, it was well-known that PARD was not a security classification. Therefore you didn't stamp PARD on something that was really secret. If it was secret RD (restricted data), it was stamped and treated as such."

    But unlike restricted data, "you could have PARD lying around in your office overnight and there was no problem," Clark said. "You could leave it in the hall and on your shelves. It was an administrative infraction to take it home."

    In time, electronic data storage made punch cards and printouts obsolete. The lab expanded PARD to embrace electronic data files.

    That, Holmer said, was never DOE's intention.

    "The intent was for hard copy," he said. "Over time, some people migrated it inappropriately to magnetic media."

    Still, under DOE rules, all data on a classified network should be considered classified until formally declassified, including the PARD downloaded by Lee, Holmer said.

    "We know it's classified because it came out of a classified computer. It is classified until it's undergone a classification review to prove it's not classified," he said.
    "That's the department's policy, and that's always been the department's policy, and that's the policy across the federal government."

    And yet, Holmer said, "Legally, we can't call it classified until it's undergone a classification review."

    But there were differences between DOE policies and practices at Los Alamos.

    During Lee's downloads, the lab assigned a lower level of computer security to PARD than to restricted data, a fact of which Holmer said he was unaware.

    A Los Alamos manual on classified computing, dated 1996 and in effect during Lee's 1997 downloads, set up a numerical ranking for data files.

    The Los Alamos computer-security hierarchy set PARD file access at security Level 5, just above unclassified data at Level 3. (No Level 4 is identified.) Confidential restricted data, the lowest classification level of restricted data, is set at Level 6. Secret restricted data is set at Level 9.

    A prosecution exhibit that tracks Lee's alleged downloads identifies all of them as PARD of Level 5.

    The Lee case could be the nation's first courtroom answer to the question of when nuclear secrets are really secret, said Aftergood of the Federation of American Scientists.

    "All of this is new," he said. "It has not been litigated before. So it has potential significance into the future."

Energy Secretary Bill Richardson distanced himself from the declassification policies of his predecessor Hazel O'Leary on a Fox network news program on April 6. "We have different philosophies on security, Hazel and I." In the brief exchange, he did not specify which categories of information he believes O'Leary should have kept classified. See: