Elon Musk shared a video clip this week of Citadel CEO Ken Griffin warning that agentic AI is now automating high-skilled finance jobs — work that used to take master's and PhD holders weeks or months. Griffin, a longtime AI skeptic, made the remarks at Stanford University and admitted the technology has changed how Citadel operates internally. The admission comes as crypto-native firms race to build product roadmaps around agents that trade and settle directly.
Griffin's Stanford confession
Griffin told students that the AI toolkit has become profoundly more powerful in just nine months. At Citadel, work typically done by quantitative researchers from top math and physics programs now takes hours or days. He said he went home one Friday 'fairly depressed' by the change. These are extraordinarily high-skilled jobs being automated, he noted — the kind Citadel has historically hired hundreds of people for.
AI and the job market
Tech employers cited AI as the trigger behind thousands of layoffs in 2026. Agentic systems account for a growing share of those cuts. Musk has long argued AI will eliminate most paid work over time. But a recent a16z review of four major studies found AI is not killing jobs at scale yet — displacement is concentrated in narrow tasks. For now, the debate is whether Griffin's Friday depression spreads across finance.
Crypto's agentic turn
Crypto-native firms aren't waiting. Coinbase, Microsoft, and other large employers frame recent cuts as a pivot toward smaller, AI-augmented teams. Several projects already let agents trade and settle on-chain without human intermediaries. The question is how much capacity Citadel and others will free up by handing PhD-level work to agents — and whether those displaced researchers will end up building the next wave of crypto protocols.
What to watch
Earnings calls this quarter could reveal which firms quantify AI-driven cost cuts. Investors will watch whether Citadel publicly details how much work it has shifted to agents. If Griffin's depression turns into a trend, the talent pool for both traditional finance and crypto could look very different by the end of 2026.
