The Trump administration on July 7, 2026, gave OpenAI the green light for a broad public launch of GPT-5.6, a source familiar with the matter confirmed to Axios. OpenAI’s flagship model Sol, plus lower tiers Terra and Luna, launch publicly Thursday.
Quick Summary – TLDR:
- OpenAI’s GPT-5.6 flagship “Sol” and its lower-cost “Terra” and “Luna” tiers launch publicly this Thursday after the government approval.
- The clearance follows additional testing and meetings between OpenAI and government officials, conducted per CAISI’s own review process inside the Center for AI Standards and Innovation..
- The Trump administration had pushed OpenAI to conduct a staggered release last month, limiting initial access to government-approved entities.
- Rival Anthropic faced a near-identical process: the Commerce Department disabled Anthropic’s Mythos 5 and Fable 5 models for all users after a June 12 export control order, then lifted the curbs last week once Anthropic added safeguards.
- Elon Musk said his company SpaceXAI is separately making its Grok 4.5 model available to the public the same week.
What Happened?
The same sequence played out for Anthropic a month earlier, and for Elon Musk at SpaceXAI this week. OpenAI’s growing footprint shows why the review now takes dedicated technical liaisons rather than a single sign-off.
OpenAI announced GPT-5.6 Sol, Terra and Luna will launch publicly this Thursday in a post on its official X account, adding: We’re expanding preview access globally now. The approval came after additional testing and meetings between OpenAI and government officials, with testing conducted by the Center for AI Standards and Innovation inside the Department of Commerce.
OpenAI sent technical experts who remained in Washington, D.C., to answer questions from officials during the review. The clearance ends a limited rollout ordered roughly 5 weeks earlier: the administration had pushed OpenAI to stagger GPT-5.6’s release last month, restricting initial access to government-approved entities. OpenAI had limited GPT-5.6 access to a small group of vetted partners, whose details it shared with authorities.
OpenAI said the staggered rollout was not its preferred way to release a new model, and said AI firms and the government are operating before more concrete standards for releasing such models, called for in President Trump’s latest AI executive order, have been finalized.
GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday.
— OpenAI (@OpenAI) July 8, 2026
We’re expanding preview access globally now. pic.twitter.com/Uk5HcfSc2e
The Anthropic Precedent
GPT-5.6’s path mirrors what Anthropic just went through. The Commerce Department banned foreigners from having access to Anthropic’s Mythos and Fable models under a June export-control order, essentially forcing their withdrawal from the market, while Anthropic had abruptly disabled the same Mythos 5 and Fable 5 models for all users following that June 12 order. The ban on Fable was lifted the prior week, with customer access restored a day later, after Anthropic implemented additional safeguards.
Both frontier labs cleared the same bottleneck within weeks of each other: a government-ordered lockdown, a technical remediation period, then a case-by-case release. OpenAI said its GPT-5.6 Sol was competitive with Anthropic’s Mythos Preview on the ExploitBench cybersecurity benchmark, the same class of dual-use capability, spanning advanced coding, biology, and cybersecurity reasoning, that triggered scrutiny of both companies’ models in the first place.
The pattern is not coincidental. Washington has increased scrutiny of advanced AI model releases to identify potential misuse by the military or the intelligence establishment in China, Russia and other countries, and Chinese authorities have separately held meetings with top domestic tech firms about potentially restricting overseas access to China’s most advanced AI models, including those yet to be released. The GPT-5.6 and Anthropic reviews look less like isolated enforcement actions and more like early runs of a repeatable US export-control lane for frontier AI, worked case-by-case rather than under a single blanket rule.
The competitive backdrop moved fast: Elon Musk, whose company SpaceXAI rivals both Anthropic and OpenAI, said his company was also making its leading model, Grok 4.5, available to the public around the same time.
Based on strong positive feedback from customers in our beta test program, @SpaceXAI will make Grok 4.5 available to the public tomorrow.
— Elon Musk (@elonmusk) July 8, 2026
It is an Opus-class model, but faster, more token-efficient and lower cost.
What’s Next?
OpenAI’s global preview access is expanding now alongside the Thursday launch. Businesses evaluating GPT-5.6 should expect continued government coordination rather than a one-time clearance, since the same staggered-then-review pattern already played out once for OpenAI and twice for Anthropic’s two models. Enterprises building on frontier models should watch for CAISI guidance updates and the concrete release standards still pending under Trump’s AI executive order; until those land, future flagship releases from any of the three labs likely face a similar case-by-case check rather than a published rulebook.
SQ Magazine’s Takeaway
This clearance reads as a second data point in a pattern, not a one-off reversal. Two competing labs, disabled or delayed over similar dual-use capability concerns, both cleared within roughly a month after adding safeguards and submitting to direct government testing. That repeatability points to the Commerce Department building a working case-by-case review process rather than improvising each time, even though the standards OpenAI itself called for during the delay still haven’t shipped.
The Grok 4.5 timing adds a competitive-pressure angle the two national-security reviews don’t fully explain. With SpaceXAI, OpenAI, and Anthropic all clearing frontier releases within days of each other, this looks as much like a market where no lab can afford to sit out a release cycle as it does a coordinated regulatory rollout.
