Citadel - The most profitable hedge fund in history finally embraced AI
What Citadel Won’t hand to a machine
The line it drew is the whole story.
Ken Griffin doesn't say things by accident. So when the founder of Citadel told an investor conference in October 2025 that generative AI, for the one job that actually matters at a hedge fund, "just falls short," people wrote it down.
Finding an edge the rest of the market hasn't found is the entire game, and the most successful trader of his generation was telling the room the machines couldn't do it.
He'd said much the same at Davos a couple of months earlier, where he treated a good deal of the AI boom as hype running ahead of proof.
Seven months later he sounded like a different man. At the Stanford Leadership Forum in May 2026, Griffin described a step change in what the tools could do.
They'd become "profoundly more powerful" than they were the year before. Work his firm used to hand teams of master's and PhD holders, the deep research that ate weeks or months, was now getting done by AI agents in hours or days.
