Moves Signal Tighter Secrecy Within C.I.A.

The New York Times
Moves Signal Tighter Secrecy Within C.I.A.
By SCOTT SHANE and MARK MAZZETTI

WASHINGTON, April 23 — The crackdown on leaks at the Central Intelligence Agency that led to the dismissal of a veteran intelligence officer last week included a highly unusual polygraph examination for the agency's independent watchdog, Inspector General John L. Helgerson, intelligence officials with knowledge of the investigation said Sunday.

The special polygraphs, which have been given to dozens of employees since January, are part of a broader effort by Porter J. Goss, the director of the C.I.A., to re-emphasize a culture of secrecy that has included a marked tightening of the review process for books and articles by former agency employees.

As the inspector general, Mr. Helgerson was the supervisor of Mary O. McCarthy, who was fired Thursday after admitting she had leaked classified information to reporters about secret C.I.A. detention centers and other subjects, agency officials said.

Mr. Goss and the C.I.A.'s deputy director, Vice Adm. Albert M. Calland III, voluntarily submitted to polygraph tests during the leak investigation to show they were willing to experience the same scrutiny they were asking other employees to undergo, agency officials said. Mr. Helgerson likewise submitted to the lie-detector test, they said.

But Mr. Helgerson's status as the independent inspector general — a post to which he was appointed by the president and from which only the president can remove him — makes his submission to a polygraph even more unusual.

L. Britt Snider, who served as inspector general from 1998 to 2001, said in an interview on Sunday night that he had not been given a polygraph in that position, though he said he was given an initial polygraph when he arrived at the agency in 1997 as special counsel to the director.

"I've never heard of it, and it's certainly unusual," Mr. Snider said. He called it "awkward" for the inspector general to be, in effect, investigated by the agency he ordinarily investigates.

But Mr. Snider and another former senior intelligence official said that it would not be improper if Mr. Helgerson had volunteered for the polygraph to set an example for others.

Reached by telephone on Sunday, Mr. Helgerson declined to comment and referred a reporter to a C.I.A. spokesman, who said he could not comment on any aspect of the leak investigation.

Further details about the inspector general's polygraph test could not be determined.

Mr. Goss has repeatedly expressed unhappiness with what he sees as the laxity of C.I.A. employees and retirees in discussing agency matters. He has taken up the cause of tightening information controls across the board, partly in response to calls from the White House, the Congressional intelligence committees and the presidential commission on weapons of mass destruction.

Mr. Helgerson's office, which investigates accusations of lapses in the ethics or performance of agency employees, has investigated some of the most serious controversies of recent years, including cases involving accusations of detainee abuse.

Since a 1989 change following the Iran-contra scandal, the C.I.A.'s internal watchdog has been confirmed by the Senate and has reported to the Congressional intelligence committees as well as to the C.I.A. director, a shift intended to assure the position's independence.

Among the subjects handled by Mr. Helgerson's office was a report completed last year that faulted senior C.I.A. officials for lapses in the failure to prevent the Sept. 11, 2001, terrorist attacks. But Mr. Goss kept the report classified and did not punish any of those named.

Former officials say the inspector general's office has also referred more than half a dozen cases of detainee abuse to the Department of Justice, but officials there have taken no action, except for a pending prosecution of one agency contract employee charged with beating an Afghan prisoner who later died.

The "single-issue" polygraphs, which are distinct from the routine polygraphs given to agency employees at least every five years, have been conducted by the C.I.A. Security Center but with close supervision from Mr. Goss's office, one official said. Like other current and former intelligence officials, he was granted anonymity to discuss classified events at the agency without fear of retribution.

For tightly "compartmented" programs like the secret detention centers, the C.I.A.'s computer system automatically limits access to the few officers who have the proper clearance to learn details of the program. The computer keeps an audit trail of which officer has looked at which documents and when they have done it, a record that would aid investigators hunting for a leaker, officials say.

The renewed emphasis on the culture of secrecy has included a tightening of the review process for books and articles by former agency employees, said Mark S. Zaid, a lawyer who represents many authors who once worked for the C.I.A.

Authors say the agency's Publications Review Board has been removing material that would easily have been approved before. While the board in the past has generally worked with retirees to make manuscripts publishable, it now more often appears to be trying to block publication, the authors say. And reprimands for violations have become more stern, including letters warning of possible Justice Department investigations.

A C.I.A. spokeswoman, Jennifer Millerwise Dyck, denied that the Publications Review Board's standards had changed.

"The only rule is that they are not allowed to have classified information in their manuscripts," Ms. Millerwise Dyck said.

But Mr. Zaid said: "There's been a fundamental shift in practice at the Publications Review Board. There's literally been a reinstitution of the 1950's attitude that what happens at C.I.A. stays at C.I.A."

Mr. Zaid said the shift in the agency's approach to publications under Mr. Goss was most clearly illustrated by its handling of a book by Thomas Waters Jr., who wrote about his experiences as a recent agency recruit.

He said the manuscript of Mr. Waters's book, titled "Class 11: Inside the CIA's First Post-9/11 Spy Class," was approved by the Publications Review Board in September 2004 with several modest changes. Mr. Waters then sold the book to Dutton, made the changes and submitted the galleys for a final review.

In February, Mr. Zaid said, the board returned the galleys with nearly half their contents marked as classified and not approved for publication. Mr. Waters, who left the agency after two years for family reasons, has sued the agency to permit publication, and the case is pending.

"What's ironic is that it's a very positive book," Mr. Zaid said. "He had a great experience and he thought this book would be a great recruiting tool."

In other cases, Mr. Zaid said, an acquaintance was recently refused permission to publish an op-ed article that drew on material from the agency's Web site. Another client's book was turned down because, the author was told, even though no single chapter was classified, the whole manuscript revealed enough information that it had to be classified. This so-called mosaic theory of classification, Mr. Zaid said, is being used more often to prevent publication.

Another former employee with long experience having publications approved agreed that reviews had become tougher. "It takes longer and there's a much more conservative approach," the former employee said, adding that he believed that some of the deletions had crossed the admittedly fuzzy boundary between protecting classified information and censoring personal opinions.

Another retiree agreed, saying he believed the agency had begun pressing authors to excise some unclassified material from manuscripts. "It's a more complex process than it used to be," he said. "Now, they question a lot more things."

Yet another agency retiree, who has in the past received warning letters from the C.I.A. after occasionally publishing articles without seeking approval, said he had recently gotten a far more strongly worded letter. This one informed him that a file had been opened to document his transgressions that could be forwarded to the Justice Department, he said.

Mr. Goss's effort to lower the profile of the agency has apparently been extended to the Web site of its Center for the Study of Intelligence, which for years has carried unclassified articles about the history and practice of spying from the in-house journal Studies in Intelligence.

Max Holland, who has written two articles for the C.I.A. journal, recently reported in The American Spectator that the online posting of unclassified excerpts from an agency review of the failure to assess Iraq's unconventional weapons accurately had been delayed for seven months. The last issue represented on the C.I.A. Web site is from mid-2005.