President Obama last week formally initiated a review of national security classification policy, directing the National Security Adviser to prepare recommendations for revising the current executive order on classification.
“My Administration is committed to operating with an unprecedented level of openness,” the President wrote on May 27. “While the Government must be able to prevent the public disclosure of information where such disclosure would compromise the privacy of American citizens, national security, or other legitimate interests, a democratic government accountable to the people must be as transparent as possible and must not withhold information for self-serving reasons or simply to avoid embarrassment.”
The President’s memo specifically identifies “overclassification” as a “problem” and invites recommendations for “greater openness and transparency.” However, both the memo’s diagnosis and its own suggested remedies are quite superficial.
For example, the memo does not address the need to revise and update the criteria for classification. And it does not require an inquiry into the role of “need to know” restrictions, if any, in a networked information environment.
The memo proposes a National Declassification Center to facilitate collaborative declassification review. This is a welcome idea that would eliminate costly, time-consuming sequential reviews of records for declassification. But without changes to classification criteria, the Center would continue to produce the same results as before, only more quickly. Thus, if the CIA thinks that fifty year old intelligence budget figures should remain classified, which it does, the mere existence of a Declassification Center would do nothing to correct such an error in judgment.
The memo also suggests restoring a presumption against classification “where there is significant doubt about the need for such classification.” This sounds fine, but because doubt (or even “significant doubt”) is an unverifiable mental state, it cannot serve as a basis for classification policy. When a similar presumption was included in past executive orders, it had no identifiable effect on agency classification practices. (The only possible role for such a provision is if the “doubter” is a different person than the “classifier.” In other words, if an independent auditor or overseer “doubted” the need for certain information to be classified, he or she might usefully be authorized to cancel its classification. But that would be a new policy measure, not a restored one.)
A response to the President’s memorandum is due within 90 days. In the same May 27 memorandum, the President also called for a review of restrictions on so-called “controlled unclassified information.”
In April 1993, President Clinton issued his own directive to undertake reforms of classification policy. Many of the questions posed in that directive are persistent and remain unresolved today, including: What types of information continue to require protection through classification in the interest of our national security? And, What steps can be taken to avoid excessive classification? See Presidential Review Directive 29, April 26, 1993.
In a paper that should be published later this month in the Yale Law and Policy Review, I attempted to identify the conditions for successful reform of government secrecy policy.
Yesterday, the U.S. Environmental Protection Agency proposed revoking its 2009 “endangerment finding” that greenhouse gases pose a substantial threat to the public. The Federation of American Scientists stands in strong opposition.
Modernizing ClinicalTrials.gov will empower patients, oncologists, and others to better understand what trials are available, where they are available, and their up-to-date eligibility criteria, using standardized search categories to make them more easily discoverable.
The Federation of American Scientists supports H.R. 4420, the Cool Corridors Act of 2025, which would reauthorize the Healthy Streets program through 2030 and seeks to increase green and other shade infrastructure in high-heat areas.
The current lack of public trust in AI risks inhibiting innovation and adoption of AI systems, meaning new methods will not be discovered and new benefits won’t be felt. A failure to uphold high standards in the technology we deploy will also place our nation at a strategic disadvantage compared to our competitors.