The challenges posed by the use of “sensitive but unclassified” control markings were examined in a comprehensive new report (pdf) from the Government Accountability Office.
“The agencies that GAO reviewed are using 56 different sensitive but unclassified designations (16 of which belong to one agency) to protect information that they deem critical to their missions — for example, sensitive law or drug enforcement information or controlled nuclear information.”
“For most designations there are no governmentwide policies or procedures that describe the basis on which an agency should assign a given designation and ensure that it will be used consistently from one agency to another. Without such policies, each agency determines what designations and associated policies to apply to the sensitive information it develops or shares. More than half the agencies reported challenges in sharing such information.”
See “Information Sharing: The Federal Government Needs to Establish Policies and Processes for Sharing Terrorism-Related and Sensitive but Unclassified Information,” March 2006 (1.8 MB PDF).
The Office of the Director of National Intelligence (ODNI) is currently coordinating an effort to standardize governmentwide procedures for the handling of “sensitive but unclassified” information.
But the ODNI rather impudently refused to cooperate with the GAO because “the review of intelligence activities is beyond the GAO’s purview,” according to Kathleen Turner of the ODNI Office of Legislative Affairs.
The Project on Government Oversight dissected the matter here. (Also flagged by Cryptome.)
See also “Report criticizes U.S. terror info sharing” by Shaun Waterman, United Press International, April 18.
After months of delay, the council tasked by President Trump to review the FEMA released its final report. Our disaster policy nerds have thoughts.
FAS and FLI partnered to build a series of convenings and reports across the intersections of artificial intelligence (AI) with biosecurity, cybersecurity, nuclear command and control, military integration, and frontier AI governance. This project brought together leaders across these areas and created a space that was rigorous, transpartisan, and solutions-oriented to approach how we should think about how AI is rapidly changing global risks.
Investment should instead be directed at sectors where American technology and innovation exist but the infrastructure to commercialize them domestically does not—and where the national security case is clear.
AI is already consequential, but its future trajectory remains contested. Policymakers should make their assumptions explicit, focus on what can be shaped rather than what can be perfectly predicted, and build institutions that can learn and respond as evidence changes.