SAN FRANCISCO: OpenAI has launched a United States-only preview of its latest artificial intelligence model series for a limited group of trusted partners, following a request from the United States government.
The company said it briefed officials on the capabilities of its new GPT-5.6 models before release and agreed to begin with a restricted rollout. The identities of selected partners have been shared with authorities.
Restricted model access
The move comes two weeks after the White House ordered OpenAI rival Anthropic to block foreign nationals from accessing its Fable 5 and Mythos 5 models, citing national security concerns.
Anthropic later shut down access to the models, saying it could not reliably enforce the restriction.
Advanced artificial intelligence models have raised concern because of their reported ability to identify software vulnerabilities that could be exploited by hackers.
President Donald Trump earlier this month signed an executive order creating a voluntary federal review of national security risks in advanced artificial intelligence models before release.
OpenAI questions process
OpenAI said it was uncomfortable with the government access process and did not believe it should become the long-term default.
The GPT-5.6 series includes three models: Sol, the flagship model; Terra, a mid-range model for everyday work; and Luna, a faster, lower-cost option.
OpenAI said the limited preview was a short-term step toward broader availability in the coming weeks.
