Today we launch a series of interviews conducted by Rachel Legow and Matt Arnold with members of the Public Interest Declassification Board (PIDB), beginning with PIDB Chair Ezra Cohen.
The PIDB is charged with advising the President and executive agencies on policy regarding classified materials, to promote public access to material. Chairman Cohen highlighted both economic and public policy benefits of maximizing transparency and access to government records, identifying these as key characteristics of the U.S. system of democratic governance – unique among nations in that even the most-guarded government secrets may eventually see the light of day. Mr. Cohen noted that over-classification and lack of transparency may inadvertently lead to false conspiracy theories (such as those surrounding the Kennedy assassination and the 9/11 terrorist attack) at great cost to societal health. Over-classification also limits government transparency, decreases U.S. ability to coordinate with allies and partners, and makes weapons and technology acquisition and development more expensive and time-consuming, all of which harm U.S. national security interests.
Mr. Cohen also pointed out the direct economic costs of over-classification, noting that storing and maintaining classified records adds significant expense and diverts resources from other needs.
Despite a broad bipartisan consensus on the need to reduce over-classification and increase transparency, Mr. Cohen notes that progress in reforming records classification standards and procedures has been slow in coming; one of the greatest challenges is overcoming bureaucratic inertia. Mr. Cohen identified two potential solutions and approaches to accelerate reform: one behavioral, and one technological.
The behavioral solution addresses the challenge of getting the bureaucracy to embrace transparency as being in its interest, by changing systemic incentives to promote, rather than discourage, appropriate levels of classification and transparency. One suggested approach would be to implement a process by which reporting over-classification leads to a reward or bonus for federal employees. Currently, penalties for under-classification (security violations, administrative penalties, even loss of job) promote risk-averse behavior, encouraging over-classification (which has no consequences). Monetary bonuses for identifying over-classification and penalties such as negative performance reviews for habitual over-classification, would support decisions to ensure that classification levels reflect the actual sensitivity of the information in question.
Effective technological solutions would apply artificial intelligence and machine learning applications to identify and propose appropriate classification levels based on a contextual algorithmic analysis of document’s content. A system-recommended classification level, with human decision-making, would both streamline and apply appropriate classification at an early stage, reducing the need for more expensive and time-consuming classification review.
Mr. Cohen pointed out that implementing either or both solutions would require resources and need to be applied and coordinated across the executive branch.
Looking forward, Mr. Cohen believes that in coming years the PIDB should focus on the priorities of (1) changing incentives and implementing technological solutions to reduce over classification in the first instance; (2) apply technological solutions to the rapidly growing holdings of historical classified records; (3) Remain responsive to additional requests for the President of the United States and Congress for assistance.