Ken Griffin: Agentic AI Replacing PhD-Level Finance Roles

At Stanford University, Citadel CEO Ken Griffin told students that agentic artificial intelligence can perform PhD-level finance work in hours. Elon Musk reposted a short clip of the remarks on his social feed.

Griffin noted the AI toolkit has become “profoundly more powerful” in about nine months. He described systems that once assisted analysts now completing work that previously took weeks or months for master’s and PhD holders in hours or days.

He described going home one Friday “fairly depressed” after seeing the technology at scale inside the firm and warned the surge in AI capability is displacing highly skilled roles. He told students, “These are not mid-tier white-collar jobs. These are extraordinarily high-skilled jobs being automated by agentic AI.”

Citadel has historically recruited hundreds of quantitative researchers from top mathematics and physics programs. Griffin’s comments indicate agentic systems are beginning to compete directly with that talent pool, handling tasks that used to require advanced degrees and long project timelines.

The Citadel example aligns with a wider trend in 2026. Technology employers have attributed thousands of layoffs to AI this year, and firms report agentic systems are an increasing factor in workforce reductions. A review by a16z of four major studies found displacement so far is concentrated in narrow tasks rather than whole occupations.

Some finance and crypto firms are building products that use agents to trade and settle transactions directly. Larger employers including Coinbase and Microsoft have described recent staff cuts as part of a shift to smaller, AI-augmented teams.

Analysts and investors will use upcoming corporate earnings reports and hiring plans to measure whether recent productivity gains persist. Market participants may also look for any public accounting from Citadel on how much capacity has been freed by assigning research-level tasks to agentic systems.