White House taps OpenAI's brakes on GPT-5.6; crypto can't stop laughing at Sol, Terra and Luna 🚀

OpenAI on Friday launched a limited preview of its GPT-5.6 family of AI models, codenamed Sol, Terra and Luna, after the Trump administration asked the company to restrict initial access to a small group of government-approved partners while federal officials evaluate the system. The rollout marks the second time this month that the U.S. government has intervened to limit the release of a frontier AI model, following its order that Anthropic suspend public access to Claude Fable 5 and Mythos 5 over national security concerns.

According to reports by The Information and Axios, the White House's Office of the National Cyber Director and Office of Science and Technology Policy asked OpenAI to limit GPT-5.6's deployment while the administration develops a framework for evaluating advanced AI models before wider release. Sources familiar with the discussions said the request was driven by GPT-5.6's "Mythos-like" capabilities rather than a broader shift in AI policy. A source told The Information that both the administration and OpenAI's developers believe GPT-5.6 is "on par" with Mythos.

OpenAI confirmed the arrangement, writing, "As part of our ongoing engagement with the U.S. government, we previewed our plans and the models' capabilities ahead of today's launch. At their request, we are starting with a limited preview for a small group of trusted partners whose participation has been shared with the government, before releasing more broadly." CEO Sam Altman told employees in a memo on Thursday that the new approval system does not represent OpenAI's long-term strategy.

The company said GPT-5.6 Sol is its flagship model, with improvements across coding, biology and cybersecurity, while Terra delivers performance comparable to GPT-5.5 at roughly 2x lower cost, and Luna is positioned as a fast, low-cost option. OpenAI introduced "max" and "ultra" reasoning modes that give Sol more time to solve complex problems or coordinate multiple subagents. On TerminalBench, a benchmark for command-line software engineering, OpenAI claimed Sol achieved the highest score and outperformed GPT-5.5, Claude Mythos 5 and Fable 5. On GeneBench v1, which evaluates long-horizon genomics and quantitative-biology analyses, the company said Sol produced stronger results than GPT-5.5 while using fewer tokens. OpenAI added that the model remains below its Cyber Critical threshold because, although it can identify vulnerabilities and exploit components, it could not autonomously produce a complete exploit chain during testing.

The model names drew immediate attention in crypto markets because they mirror the tickers of established digital assets. Solana's native token trades as $SOL, while Terra and Luna were the names behind the Terra blockchain ecosystem whose 2022 collapse wiped out tens of billions of dollars in market value. OpenAI stated that the names are not linked to cryptocurrencies but connote different tiers of capability, with Luna positioned as the entry-level option. The selection nonetheless revived some of the most recognizable brands in the digital asset space and prompted social media users to flag the resemblance within hours of the announcement.

The administration's intervention follows an executive order earlier this month directing federal agencies to establish a voluntary testing framework for advanced AI systems before release. OpenAI, Anthropic and Google have each published proposals outlining how frontier AI should be governed, with all three calling for structured evaluations of the most capable models, greater transparency around safety testing, independent review of high-risk systems, and a larger role for the government in overseeing AI development.