DOD Provides Update on AI-Powered Record Declassification Initiative

The Department of Defense has provided an update on a research initiative exploring the viability of artificial intelligence and machine learning in managing Pentagon documents.

According to J.D. Smith, chief of the records and declassification unit at the Pentagon’s Washington Headquarters Services, the research effort has validated a proof of concept using AI and machine learning models to carry out records management and classify and declassify documents.

Led by the University of Maryland’s Applied Research Laboratory for Intelligence and Security, the Office of the Under Secretary of Defense for Intelligence and Security-sponsored study assesses the effectiveness of AI in helping declassification operations amid growing digital record use, Federal News Network reported.

Smith disclosed that findings from the study will help the DOD develop a “playbook” that will guide the government in using emerging technologies in declassification and record management initiatives.

The update comes as a group of bipartisan lawmakers wrote a letter to the office of federal Chief Information Officer Clare Martorana asking for the research’s progress.

Records management is not the only AI use case at the Defense Department. The Pentagon also sees potential in using it to improve moving-target intelligence. In March, it tapped space-based intelligence provider BlackSky Technology to provide its Spectra tasking and analytics platform to collect and annotate multiframe burst imagery generated from space.