On July 23, 2025, U.S. President Donald Trump takes an AI-related executive order after signing it during the “Winning the Race to AI” summit in Washington, DC.
Kent Nishimura | Reuters
Chris Fall, director of the Center for AI Standards and Innovation, has resigned from his position as director, just three months after being selected for the position by the Trump administration, CNBC confirmed on Monday.
CAISI, part of the U.S. Department of Commerce, aims to help governments facilitate “testing and collaborative research” on commercial artificial intelligence systems, according to its website.
Commerce Department spokeswoman Kristen Eichamer told CNBC that Arvind Raman, director of the National Institute of Standards and Technology, will serve as CAISI’s acting director and continue to oversee it. Eichamer did not say why Fall, who was appointed to run the organization in April, resigned.
Axios first reported Fall’s departure.
His departure adds to the uncertainty surrounding who will be at the forefront of the Trump administration’s AI policy and research. Venture capitalist David Sachs previously held the position of AI and cryptocurrency czar in the White House, but he stepped down from that role in March and his replacement has not yet been named.
This is a particularly sensitive time for the White House, as China’s open source model is gaining traction in the US versus the big proprietary models of OpenAI and Anthropic. Late last week, Chinese startup Moonshot AI announced a new model, the Kimi K3, that it claims has closed the gap with the American flagship and outperformed OpenAI and Anthropic’s most capable systems on several benchmarks.
The Trump administration has taken a more pragmatic approach to AI regulation since the president signed the AI Executive Order in June. The order requires AI developers to voluntarily provide their models to the government to evaluate their capabilities prior to full release, and federal agencies have 60 days to develop evaluation frameworks.
For companies looking to introduce powerful new models in the meantime, the process is unclear.
OpenAI announced in June that it had agreed to limit the deployment of its GPT-5.6 model series to a group of “trusted partners” at the request of the U.S. government. Two weeks ago, Anthropic had to disable access to its Fable 5 and Mythos 5 models to comply with an export control directive issued by the Department of Commerce. Both companies subsequently managed to release the model more widely.
The Trump administration has taken steps in recent weeks to implement the executive order, including announcing the creation of a clearinghouse called Gold Eagle. As CNBC previously reported, the clearinghouse is aimed at finding and fixing cyber vulnerabilities and has given the White House permission to give companies access to cutting-edge AI models.
The clearinghouse has already begun “incorporating and prioritizing identified cybersecurity vulnerabilities” and “coordinating scan verification,” according to a White House release. CAISI is not named as an institution involved in its development.
Watch: Fmr. WH Senior Policy Advisor talks about Trump administration lifting restrictions on Anthropic’s AI models
