The Central Intelligence Agency (CIA) aims to embed specialized generative AI assistants into its core analytics tools within two years, enabling officers to track foreign spies and anticipate hostile actions from abroad. Deputy Director Michael Ellis outlined the plan at a Special Competitive Studies Project event in Washington, D.C., describing the assistants as a "classified version of generative AI" that will handle routine tasks while humans retain control over key decisions.
Classified AI Assistants Roll Out in Two Years
"Within the next couple of years, we will have AI co-workers built into all of the agency's analytic platforms," Ellis said. These digital helpers will draft assessments and identify patterns in global intelligence. He stressed that human officers will remain the final authority on critical judgments. The initiative comes as the federal government's partnership with Anthropic collapses: President Trump ordered agencies to stop using the company's Claude AI in March, and the Defense Department labeled Anthropic a supply chain risk—a move the firm is challenging in court. Ellis did not name the company but noted the agency cannot afford to let "the whims of a single company constrain our capabilities."
Blockchain Data Already in Use
Ellis previously disclosed in May that the CIA tracks blockchain data to support counterintelligence, viewing cryptocurrency as a vital element in the technological race against China. "Five to ten years ago, China was nowhere near America in terms of technological innovation. That's just not true today," he said. The agency is now combining on-chain monitoring with AI analysis to sharpen its edge. The push reflects a broader urgency: the US lead is shrinking, and CIA must adopt frontier tools—from crypto tracking to automated intelligence processing—to stay ahead.