White House tightens control over access to frontier AI models

Traders Union
All News
Financial News
White House tightens AI control

Highlights

The White House is moving to control which companies and entities can access frontier AI models, overriding the previous developer-led access system.
The administration temporarily blocked Anthropic's Claude Mythos 5 and Fable 5 models in June over national security issues and prompted OpenAI to restrict new model access to trusted partners.
China's Moonshot AI launched the Kimi K3 model on Friday, rivaling U.S. models Fable and GPT-5.6 and outperforming them on at least one independent benchmark.

Government role grows in AI rollout decisions

As first reported by CNBC, the administration is taking steps to decide which companies and entities can gain access to the latest frontier AI models, a role that had largely been left to model developers. Until now, companies such as Anthropic and OpenAI have determined which agencies, businesses and enterprise customers could use their most advanced systems.

Anthropic has unveiled its cybersecurity-focused Mythos model to a limited group of partners through Project Glasswing, while OpenAI has operated a similar consortium called Daybreak for its cybersecurity model. OpenAI was also asked by the administration to gate its recent GPT-5.6 release, according to the report.

A White House official told CNBC that the government does not provide approvals for private-company AI releases and said any testing, meetings or engagements with government experts are voluntary. The official added that decisions on the timing and scope of releases rest with the companies, pointing to President Donald Trump's June executive order.

At the same time, the administration blocked Claude Mythos 5 and Fable 5 last month over national security concerns before restoring access after negotiations with Anthropic. OpenAI also said last month that it would restrict new AI models to trusted partners to comply with government requests.

Cybersecurity concerns and China competition shape policy

This week, the administration launched a program called Gold Eagle to work with the private sector on identifying and fixing cyber vulnerabilities. According to a source familiar with the matter, the clearinghouse is designed to put the White House in charge of greenlighting which companies can access new AI models.

The policy shift leaves the future of company-led access programs such as Project Glasswing and Daybreak uncertain. One source said future rollouts will require explicit government approval on which partners can participate.

The White House is trying to balance tighter oversight with concern that U.S. firms could lose ground in the global AI race. Chinese startup Moonshot AI unveiled its Kimi K3 model on Friday, a system that the report said largely catches up with Fable and GPT-5.6 and outperforms U.S. frontier models on at least one independent benchmark.

David Sacks, founder of Craft Ventures and former White House AI czar, called that breakthrough concerning. He wrote that the rest of the world will not follow the same rules if the U.S. slows itself down.

In our earlier roundup of major venture funding deals, we noted that investor appetite for AI remained strong, led by Fireworks AI’s $1.505 billion Series D and additional large rounds across AI infrastructure, life sciences AI, and robotics. The piece also highlighted how capital is flowing not only into core AI platforms but into sector-specific applications, underscoring the rapid scaling of the ecosystem that today’s policy decisions are increasingly shaping.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.