CIA to deploy “AI co-workers” in analytics to track spies and predict hostile moves

The CIA plans to embed classified generative AI “AI co-workers” across its analytic infrastructure within two years. The goal is to help analysts identify foreign intelligence trends, draft reports, and spot patterns that could signal hostile actions—while keeping humans in control of “key decisions.” CIA Deputy Director Michael Ellis said these “AI co-workers” will become standard across CIA analytic platforms, handling routine tasks and accelerating threat assessment. The initiative comes as U.S. officials tighten AI technology use following a government-wide ban on Anthropic tools and a legal dispute over Anthropic’s supply-chain risk status. Ellis also argued the CIA must remain independent of private-sector constraints. Ellis linked the push for better intelligence technology to U.S.-China competition, saying the American tech lead has narrowed. He previously noted the CIA tracks blockchain data for counterintelligence, framing cryptocurrency as part of the technological race. No crypto-specific policy change is announced, but the mention of blockchain analytics reinforces ongoing intelligence interest in crypto-linked data sources.