OpenAI is set to publicly launch GPT-5.6, its most advanced AI model series, on Thursday after the Trump administration approved a broad rollout following additional testing and government consultations.

The launch — which includes three tiers: Sol (flagship), Terra (mid-range), and Luna (cost-efficient) — was delayed last month after the U.S. government requested heightened scrutiny over national security concerns about the potential misuse of powerful AI technologies.

According to Axios, which broke the story, the White House and Commerce Department signed off after OpenAI demonstrated additional safeguards and held extensive meetings with government officials. The approval signals a shift toward a more structured framework for frontier AI releases — the same week China signaled it may restrict overseas access to its own top models.

OpenAI had initially limited GPT-5.6 preview access to a small group of vetted partners whose details were shared with authorities. The company touted improved agentic capabilities across coding, biology, and cybersecurity — noting that GPT-5.6 Sol was competitive with Anthropic's Mythos Preview on the ExploitBench cybersecurity benchmark.

The rollout comes amid an escalating U.S.-China AI race and just days after Anthropic's own advanced models were briefly disabled under a June 12 export control order before being reinstated with new safeguards. Elon Musk's SpaceXAI also announced it would make Grok 4.5 publicly available.

President Trump has signed an executive order establishing a voluntary framework under which AI developers can provide frontier models to the U.S. government for up to 30 days before releasing them to trusted partners — a policy that effectively shaped the GPT-5.6 rollout timeline.

OpenAI confirmed the launch in an X post late Tuesday, saying GPT-5.6 will be available globally starting Thursday.

Sources