NSA Declassifies Secret Document After Publishing It

05.14.12 | 5 min read | Text by Steven Aftergood

The National Security Agency last week invoked a rarely-used authority in order to declassify a classified document that was mistakenly posted on the NSA website with all of its classified passages intact.

The article is a historical study entitled Maybe You Had to Be There: The SIGINT on Thirteen Soviet Shootdowns of U.S. Reconnaissance Aircraft.  It was written by Michael L. Peterson and was originally published in the classified journal Cryptologic Quarterly in 1993.

Late in the afternoon of May 11 (not May 9 as stated on the NSA website), the NSA published a formally declassified version of the article with the annotation “Declassified and approved for release by NSA… pursuant to E.O. 13526 section 3.1(d)….”

Section 3.1(d) of executive order 13526 permits the declassification of properly classified information when there is an overriding public interest in doing so.  It is almost never cited and it is hard to think of another occasion when it has been used by any government agency to justify declassification.  It reads:

“3.1(d) It is presumed that information that continues to meet the classification requirements under this order requires continued protection.  In some exceptional cases, however, the need to protect such information may be outweighed by the public interest in disclosure of the information, and in these cases the information should be declassified.  When such questions arise, they shall be referred to the agency head or the senior agency official.  That official will determine, as an exercise of discretion, whether the public interest in disclosure outweighs the damage to the national security that might reasonably be expected from disclosure….”

So what was “exceptional” about this particular NSA historical study?  What was the overriding public interest in it that justified its complete declassification despite its presumed eligibility for continued classification?  What unavoidable damage was expected to result from its disclosure?  The NSA Public Affairs Office refused to answer these questions, despite repeated inquiries.

In fact, NSA was being disingenuous by invoking section 3.1(d).  There was nothing exceptional about the contents of the document, and there was no overriding public interest that would have compelled its disclosure if it had been properly classified.  Nor is any national security damage likely to follow its release.

Rather, the hasty NSA declassification action was intended to conceal the fact that NSA had mistakenly published the full classified text of the document on its website two days earlier, after having rebuffed regular requests for declassification.

In response to a May 2009 Mandatory Declassification Review request from aerospace writer Peter Pesavento, NSA had previously released a heavily redacted version of the article.  Mr. Pesavento appealed the case to the Interagency Security Classification Appeals Panel, and last month the Panel agreed that some additional portions of the document could be declassified, while the rest should remain classified.  The partially declassified document was still working its way through the appeal system and had still not been provided to Mr. Pesavento.

But then on May 9, the National Security Agency inexplicably published the entire document on its website.  Instead of censoring the text by blacking out the classified portions, those portions were actually highlighted, leaving the document fully available to startled readers.  After we contacted the NSA on May 10 to inquire about the classification status of the document, it was immediately removed from the NSA site.

But we retained a copy of the uncensored classified article as published by the NSA, which is available here.

Secrecy News submitted several questions to NSA Public Affairs last week about the classified document, and we indicated our intention to publish it ourselves since it did not appear to meet current classification standards.  NSA officials asked for a four-day extension of our deadline to give them time to respond to our questions, and we agreed.  But that proved to be a futile gesture on our part, since the NSA Public Affairs Office in the end refused to answer any of the questions we posed.  In retrospect, it appears that NSA never intended to answer any of our questions but simply wanted to preempt the reposting of the classified document by hastily declassifying it.

The newly disclosed article was originally classified SECRET SPOKE.  SPOKE is a now-defunct classification compartment for communications intelligence, explained intelligence historian Jeffrey Richelson, who first spotted the uncensored NSA publication online.  (It so happens that Dr. Richelson’s own work is cited in the article.)

In the classified version of the article that was posted online by NSA, all of the classified paragraphs of the article were marked with the basis for their classification.  In most cases, this was section 1.4(c) of the executive order on classification, which pertains to “intelligence activities, intelligence sources or methods, or cryptology.”  In some cases, the basis for classification was section 1.4(d) on “foreign relations or foreign activities of the United States, including confidential sources.’  In a couple of other cases, the justification cited was Public Law 86-36, which is the National Security Agency Act, a statutory non-disclosure provision.  A week ago, this material purportedly posed a threat of “serious damage” to national security if disclosed.  Now all of it has been made public.

While communications intelligence is among the most sensitive categories of national security information, this article is clearly remote from any contemporary security issues.  It reviews the record of signals intelligence coverage of thirteen episodes in which Soviet forces shot down U.S. aircraft.  But those incidents occurred between 1950 and 1964 — or many generations ago in terms of intelligence technology and practice.

On the other hand, the article does present what appears to be some valuable “new” information including some fine details about SIGINT coverage of the U-2 incident in May 1960.

But the author himself acknowledged that all of this is ancient history.

“Looking back over forty years,” he wrote in the conclusion of his 1993 paper, “it may be difficult to give sufficient weight to the level of anxiety over and ignorance about the Soviet Union experienced by Americans.  Moreover, the fear of another Pearl Harbor was very real.  The airborne reconnaissance program helped reduce these fears by erasing the ignorance.”

“Little of this concern prevails today,” he noted.  “Why all the fuss? Maybe you had to be there.”

But even “being there” does not help one to understand the erratic NSA classification practices reflected in this case.  NSA classification policy seems to be completely untethered from contemporary national security threats.

Among other things, the NSA’s abrupt declassification of the the document shows that the Interagency Security Classification Appeals Panel needs to recalibrate its document review procedures.  It is now clear that the Panel was unduly deferential to NSA, and that it erred last month by giving credence to the NSA’s claims that portions of the document warranted continued classification.  Today, not even the NSA says that.