Third Leader Out in Four Months at US AI Standards Agency as Chris Fall Resigns from CAISI

Chris Fall has resigned as director of the Center for AI Standards and Innovation (CAISI) after just three months in the job, making him the third person to exit leadership of the agency since the start of the year. No official reason was given for his departure.

Fall's brief tenure followed a chaotic run at the top of CAISI, which operates inside the National Institute of Standards and Technology and is tasked with building technical standards, evaluation frameworks, and cybersecurity risk assessments for AI systems. He took over after Collin Burns lasted less than a week in the role, reportedly forced out over his prior work at Anthropic amid friction between the Trump administration and the company. Before that, David Sacks, the White House's AI and crypto czar, had stepped away from leading CAISI in March 2026. Fall himself came from a background in federal science administration, having previously directed the Department of Energy's Office of Science and held posts at ARPA-E and the Office of Naval Research.

The revolving door at CAISI comes at a moment when the administration's approach to AI oversight is under scrutiny from multiple directions. In June, the Commerce Department invoked export control authority to force Anthropic to temporarily withdraw its Mythos and Fable models from the market, only to lift the ban weeks later after Commerce Secretary Howard Lutnick said he was satisfied with the company's safety measures. Around the same period, Chinese lab Moonshot released an upgraded version of its open-weight Kimi model that matched flagship systems on performance benchmarks, reportedly triggering internal administration debate over whether to restrict Chinese open models in the US market.

Notably, CAISI itself was left out of the White House's newest AI oversight initiative. The administration signed an executive order establishing "Gold Eagle," a program creating a clearinghouse for coordinating AI-related cybersecurity vulnerabilities, but assigned responsibility to the Commerce Department and the Department of Homeland Security rather than CAISI, the agency ostensibly built for exactly this kind of technical risk work.

The instability has not gone unnoticed in the industry. Google DeepMind CEO Demis Hassabis has publicly argued for a different model altogether: an independent, industry-funded standards body along the lines of FINRA, the self-regulatory organization that oversees broker-dealers, rather than a shifting cast of political appointees inside a federal agency.

For an administration that has positioned AI competitiveness and safety oversight as a policy priority, the churn at CAISI raises questions about institutional continuity just as frontier model capabilities, and geopolitical competition around them, are accelerating. Three directors in roughly six months leaves the agency without the sustained leadership typically needed to set durable technical standards, even as new flashpoints, from export bans to open-weight rivals from China, keep landing on its doorstep.