OpenAI is releasing GPT-5.6 globally today after the U.S. Department of Commerce cleared the way for a broad rollout of its most advanced artificial intelligence model. The company will launch three versions—GPT-5.6 Sol, Terra, and Luna—following a month-long review process with federal officials focused on national security risks posed by powerful AI systems.
The rollout marks the end of a deliberate pause that began when the Trump administration requested OpenAI stagger its release about a month ago. That initial request reflected growing concern in Washington about how advanced AI could be weaponized by military or intelligence agencies in China, Russia, and other countries. The administration didn't formally grant permission—a White House official emphasized late Wednesday that "no such permission is required or granted"—but the company worked closely with government officials on safety testing before proceeding.
How the Review Process Worked
OpenAI CEO Sam Altman visited the White House and lawmakers about a month ago, beginning what became an intensive vetting period. The company then spent more than a month previewing and testing GPT-5.6 with the White House and the Center for AI Standards and Innovation. On June 26, two weeks ago, OpenAI announced it would comply with the administration's request for a staggered release, saying the approach would "ensure the model can be distributed widely in the coming weeks."
The company's own statement revealed a tension at the heart of the process. While agreeing to the temporary pause, OpenAI wrote: "We don't believe this kind of government access process should become the long-term default." The statement framed the review as a "short-term step" necessary to develop what it called a "repeatable process for future model releases" under the administration's cyber executive order framework.
An OpenAI spokesperson confirmed that government approval wasn't technically necessary, but that the company was "actively working with the administration on safety testing and reviews before deploying GPT-5.6 widely." The spokesperson added that the White House signaled comfort with the plan once that testing concluded.
The Competitive Pressure
The release comes as AI developers race to improve performance and cut costs. OpenAI said GPT-5.6 Sol is competitive with Anthropic's Mythos Preview on the ExploitBench cybersecurity benchmark. Billionaire Elon Musk, whose company SpaceXAI competes with both Anthropic and OpenAI, announced Wednesday that his Grok 4.5 model would become publicly available.
But there's a larger competitive dynamic at play. Chinese AI developers are reshaping the economics of the sector by delivering increasingly capable models at a fraction of the cost. Zhipu, trading as Knowledge Atlas Technology JSC, launched its GLM 5.2 model last month and made it free to download, fine-tune, and run on enterprise servers. According to reporting from CNBC, the U.S. government's tight grip on domestic frontier AI is creating an unintended opportunity for Chinese competitors, who are leveraging the pause to gain ground with more accessible, cost-effective models.
Why This Matters:
The GPT-5.6 rollout represents a working model of how democratic governments can exercise oversight over transformative technologies without blocking innovation entirely. The month-long review process allowed federal officials to assess national security risks while OpenAI maintained the ability to release its product. However, the process also reveals structural challenges: OpenAI's resistance to making government review a permanent requirement suggests ongoing tension between corporate autonomy and public accountability. Meanwhile, the competitive advantage gained by Chinese developers during the pause illustrates how regulatory caution in one country can shift global market dynamics. As AI systems become more powerful and more widely deployed, questions about who sets the rules for their release—and how those rules are enforced—will only intensify. The voluntary nature of OpenAI's cooperation with the administration underscores that this framework relies on corporate goodwill rather than statutory authority.