The Trump administration is increasingly dictating which companies and entities can access the latest frontier AI models, shifting control away from tech giants like Anthropic and OpenAI, according to two people familiar with the matter.
Until recently, Anthropic and OpenAI decided which partners received their most powerful models, often prioritizing major enterprise customers. Anthropic's Mythos cybersecurity model was shared with a select group through Project Glasswing, while OpenAI's GPT-5.6 release was gated at the administration's request, with a similar consortium called Daybreak for cybersecurity.
Government oversight expands
A White House official told CNBC that the administration does not approve AI releases from private companies, calling any engagements "voluntary" and noting that "decisions on timing and scope of releases rest entirely with the companies." The official pointed to President Donald Trump's June executive order, which asked companies to voluntarily provide early access for testing.
However, last month the administration blocked Anthropic's Claude Mythos 5 and Fable 5 over "national security concerns," restoring access only after weeks of intense negotiations. OpenAI also said it would limit new models to "trusted partners" to comply with government requests.
New program and competitive pressures
This week, the administration launched "Gold Eagle," a program to collaborate with the private sector on finding and fixing cyber vulnerabilities. A proposed "clearinghouse" would put the White House in charge of approving which companies can access new AI models, according to a person familiar with the matter.
The moves cast doubt on company-led initiatives like Project Glasswing and Daybreak. Going forward, these rollouts will require explicit government approval for partners, one source said.
Meanwhile, Chinese startup Moonshot AI unveiled its Kimi K3 model on Friday, which largely matched the performance of Fable and GPT-5.6 and even outperformed them in at least one independent benchmark. David Sacks, former White House AI czar, called the breakthrough "concerning," warning that overregulation could cede the AI race to competitors.