The Central Intelligence Agency said this week that it will post its database of declassified CIA documents online, making them broadly accessible to all interested users.
The database, known as CREST (for CIA Records Search Tool), contains more than 11 million pages of historical Agency records that have already been declassified and approved for public release.
Currently, however, CREST can only be accessed through computer terminals at the National Archives in College Park, MD. This geographic restriction on availability has been a source of frustration and bafflement to researchers ever since the digital collection was established in 2000. (See CIA’s CREST Leaves Cavity in Public Domain, Secrecy News, April 6, 2009; Inside the CIA’s (Sort of) Secret Document Stash, Mother Jones, April 3, 2009).
But that is finally going to change.
The entire contents of the CREST system will be transferred to the CIA website, said CIA spokesperson Ryan Trapani on Tuesday.
“When loaded on the website they will be full-text searchable and have the same features currently available on the CREST system at NARA,” he said.
CIA was not able to provide a date for completion of the transfer, but “we are moving out on the plan to make the transition,” Mr. Trapani said.
In the meantime, “The CREST database housed at NARA will remain up and running at least until the website is fully functioning,” he said.
Protecting the health and safety of the American public and ensuring that the public has the opportunity to participate in the federal decision-making process is crucial. As currently organized, FACs are not equipped to provide the best evidence-based advice.
As new waves of AI technologies continue to enter the public sector, touching a breadth of services critical to the welfare of the American people, this center of excellence will help maintain high standards for responsible public sector AI for decades to come.
The Federation of American Scientists supports the Critical Materials Future Act and the Unearth Innovation Act.
By creating a reliable, user-friendly framework for surfacing provenance, NIST would empower readers to better discern the trustworthiness of the text they encounter, thereby helping to counteract the risks posed by deceptive AI-generated content.