Navy Considers Using Generative AI in Automating Security Classification

The U.S. Navy is considering the use of generative artificial intelligence to automate the assignment of security classifications. 

David Broyles, research program director at the Center for Naval Analysis, said at the DON IT West conference on Wednesday that analysts can depend on large language models to analyze datasets to gain a better understanding of the implications of security classification and to more accurately assign classification labels. 

The decision to consider using the technology comes amid the growing volume of restricted documents, Federal News Network reported.

Bill Streilein, chief technology officer at the Office of the Chief Digital and AI Officer, said modeling and simulation help address the challenges associated with working with classified data. 

The use of AI is not new to the service branch. In May 2023, Navy Secretary Carlos Del Toro, a Wash100 Award winner, disclosed that the Navy is working on AI-enabled unmanned systems to help increase maritime domain awareness in a bid to outpace the nation’s adversaries. 

In February, Vice Adm. Craig Clapperton, head of the U.S. Fleet Cyber Command and the 10th Fleet, said the service is turning to automation and data analytics to improve its cyber posture and reduce cyberthreats.