The Trump administration reportedly does not plan to release its highly anticipated artificial intelligence oversight framework to the public after presenting it to technology giants.
The White House on Tuesday held a meeting with AI industry partners to discuss a completed framework to test certain AI models’ capabilities. The White House does not plan to release its framework to the public, according to Axios, citing three sources familiar with the plans. It remains unclear which “trusted partners” would also have access to frontier AI models before the frontier AI models’ public release.
President Donald Trump wrote an executive order in June calling for the creation of AI safety testing framework.
The White House did not immediately respond to a Daily Caller News Foundation request for comment, citing sources familiar with the matter.
The administration’s AI testing plan will exclude open-source AI models from testing, Axios reported.
Nvidia and other developers of open-source models might have to submit their models as they become more advanced, the Wall Street Journal reported, referencing people familiar with the matter. Nvidia CEO Jensen Huang recently discussed AI oversight with the White House. The company launched the Open Secure AI Alliance, which includes many leading tech companies such as Palantir, IBM, Crowdstrike and SpaceX, to advocate for open-source AI models.
The White House framework’s exclusion of AI models from its framework follows as the Trump administration appeared more interested in restricting access to Chinese open-source AI Models after China closed the gap with its latest models.
Chinese lab Moonshot AI in July released its latest model Kimi K3 which now threatens America’s leadership on the technology. The White House increased its scrutiny of Chinese open-source AI as its Trump administration officials said they believe that Moonshot AI used Anthropic’s flagship AI model to train Kimi K3 through a process known as “distilling.”
(DCNF)