CIA's New Work Buddies: AI 'Coworkers' Join the Spy-Hunting Squad

The CIA is getting ready to embed AI co-workers directly into its analytics platforms to help analysts sniff out spies and anticipate hostile moves by foreign adversaries. Basically, Skynet but with a security clearance and a 401(k).

Within the next couple of years, the agency will have AI built into all of its analytic platforms—basically a classified version of generative AI to help with the grunt work, according to CIA deputy director Michael Ellis, who made the announcement Thursday at an event hosted by the Special Competitive Studies Project in Washington, DC. Think ChatGPT but it can't tweet about being sentient and instead just quietly judges whether that foreign diplomat is actually a spy. Fun times.

These digital coworkers will assist intelligence officers with drafting key judgments, testing analytical conclusions, and identifying trends in intelligence gathered from abroad. But don't worry—humans will still be the ones pulling the trigger on the big decisions. The robots get to do the busywork, while actual humans get to keep the existential responsibility. Someone has to sign off on regime changes at 3 AM.

The CIA has already been kicking the tires on AI, having tested roughly 300 AI projects last year to bring new capabilities to its mission, including processing large data sets and language translation. The agency even recently produced its first entirely AI-generated intelligence report. That's right, folks—some analyst's coffee break is now fully automated. The machine learned, and now it's writing briefings. We truly are living in the future.

Ellis also made sure to note that the US can't let a single company's whims constrain its capabilities—a not-so-subtle nod to the ongoing feud between the Department of Defense and AI firm Anthropic. Despite Anthropic's $200 million contract with DoD, the company refused to let its flagship AI product Claude be used for mass domestic surveillance or fully autonomous weapons. President Trump ordered all federal agencies to stop using Anthropic's tech in March, and the legal wrangling continues. Nothing says "trust the technology" like a multi-hundred-million-dollar contract that ends in a messy breakup and a lawsuit. Classic.

Staying ahead of China is also a major driver here. Ellis pointed out that the technological gap between the US and China has narrowed significantly—five to ten years ago, China was nowhere near America's level in tech innovation, but that's no longer the case. The CCP went from "let's copy that iPhone" to "let's build our own AI that doesn't ask uncomfortable questions" in what feels like a blink. Yikes.

Oh, and in case you were wondering: Ellis is also pro-transparency when it comes to Bitcoin. Back in May, he said crypto and Bitcoin are matters of national security, and the agency actually uses blockchain data to assist with its counterintelligence operations. Because apparently, even spies need to track on-chain movement these days. Nothing like following the money—except now the money is a public ledger that never forgets. Sats over secrets, as they say.