When AI Gets Too Good,
Spies Go Old School
As a spy thriller author, I spend a lot of time thinking about what real intelligence operations look like — not the Hollywood version, but the actual mechanics of tradecraft. So when the CIA's own academic journal publishes something that reads like a plot twist, I pay attention.
A recent analysis in Studies in Intelligence — the CIA's in-house journal, not light bedtime reading — makes a case that would have sounded absurd five years ago: artificial intelligence has become so effective at fabricating reality that spies may have no choice but to go back to meeting in person.
Dead drops. Brush passes. Face-to-face contact in carefully chosen locations. The oldest tools in the tradecraft handbook — the ones that Blake MacKay relies on — are suddenly looking like the most trustworthy options intelligence professionals have.
"The same technologies enhancing intelligence gathering may ironically make it harder to trust the data those tools produce."
— Nextgov/FCW, citing CIA Studies in Intelligence, April 2026The logic is sound. If an adversary can generate a convincing deepfake video of a senior official, fabricate a voice call, or manufacture a chain of authentic-looking digital communications — then every piece of digital intelligence becomes suspect. The only thing that can't easily be faked is two people in the same room.
The irony cuts deep: the technology that makes intelligence agencies more powerful is simultaneously making their work harder to trust. The CIA paper isn't calling for a wholesale retreat from AI — it's raising a harder question about what happens to signal intelligence when the signal itself can be forged.
Key Takeaways
- AI-generated deepfakes and fabricated communications are eroding trust in digital intelligence sources
- Classic HUMINT techniques — dead drops, brush passes, in-person meetings — may be regaining strategic value
- The CIA's own analysts are publishing on this tension between AI capability and data integrity
- Intelligence services may be forced to run parallel tracks: AI for scale, human contact for verification
This isn't science fiction. A federal technology publication doesn't cite CIA journal articles about hypothetical problems. The people who run real intelligence operations are wrestling with this right now.
And it's exactly the world Blake MacKay operates in — where the technology is dazzling, the data can't always be trusted, and sometimes the only way to know something for certain is to put a human being in the room.
The Blake MacKay Connection
Blake MacKay's tradecraft has always relied on a combination of cutting-edge intelligence and old-school instinct. Start the series free with Intercept — and see how the world this article describes plays out at street level, in real time, with lethal stakes.
Read the Fiction
Behind the Intelligence
CIA operative Blake MacKay lives in the same world these briefings describe. Start with the free prequel and see how real tradecraft becomes high-stakes fiction.
Get Intercept Free →