US AI Safety Institute Head Chris Voigt Resigns After Three Months
Chris Voigt, the head of the US Center for AI Standards and Innovation (formerly the US AI Safety Institute), has resigned after only three months in office. Arvind Raman, the head of the Commerce Department office managing the center, will temporarily step in to lead the organization. This sudden leadership change highlights the ongoing instability and shifting priorities in US government AI policy under the Trump administration. It could impact the government's collaborative safety testing agreements with major AI labs like OpenAI, Anthropic, and Google DeepMind. The center is tasked with evaluating advanced AI models for "demonstrable risks," such as assisting adversaries in developing chemical or biological weapons. Voigt's departure comes amid reports that the institute's regulatory and oversight capabilities may face significant policy shifts or reductions.
## BACKGROUND
The US AI Safety Institute was established to evaluate and ensure the safety of frontier AI models, partnering with major developers on safety research. In 2025, it was renamed the Center for AI Standards and Innovation (CAISI). Leading AI labs like OpenAI and Anthropic have recently raised concerns about advanced models potentially assisting in the creation of chemical and biological threats, prompting stricter safety protocols.