The Central Intelligence Agency plans to integrate AI “co-workers” into its analytics platforms to help process intelligence, detect spies, and anticipate threats. Deputy director Michael Ellis said the tools will assist with drafting assessments, testing conclusions, spotting trends, and translating large datasets, while humans will continue making final decisions.

The initiative follows the agency’s testing of around 300 AI projects, including data analysis and automated report generation. The rollout also comes amid tensions between the US Department of Defense and Anthropic, after the firm restricted use of its AI systems for surveillance and autonomous weapons, leading to a broader dispute over national security technology access.

Officials say the push is partly driven by competition with China, as the US seeks to maintain an edge in AI-powered intelligence. Ellis also highlighted the transparency of Bitcoin and blockchain data, noting that crypto analytics are increasingly used in counterintelligence operations.