OpenAI will release its most powerful artificial intelligence model yet on Thursday, marking the first major rollout of a frontier AI system under the Trump administration's new hands-on regulatory framework. The U.S. Department of Commerce cleared the way for the broad release of GPT-5.6 after the company conducted additional testing and met with government officials, according to Axios and Reuters reporting.
The three-tier rollout—featuring the flagship Sol model alongside lower-tier Terra and Luna versions—comes after less than one month of heightened government scrutiny. The delay itself signals a fundamental shift in how Washington approaches powerful AI technology: no longer hands-off, but actively assessing capabilities before public release.
The Regulatory Pivot
President Trump signed an executive order establishing a voluntary framework requiring AI developers to submit "covered frontier models" to the U.S. government for up to 30 days before releasing them to trusted partners. OpenAI initially resisted this approach. The company stated at the time that a staggered rollout was not its preferred way to release new models, and noted that AI firms and government are operating before more concrete standards for releasing such models have been finalized.
The Center for AI Standards and Innovation within the Department of Commerce conducted the testing, with OpenAI sending technical experts who remained in Washington to address questions. OpenAI had limited GPT-5.6 access to a small group of vetted partners whose details were shared with authorities, though the company didn't disclose the names of these partners.
National Security and Global Competition
Washington has intensified scrutiny of advanced AI releases over concerns the technology could be misused by military or intelligence agencies in China, Russia, and other countries. Reuters reported that the United States and China are engaged in a race to develop cutting-edge AI models that could dramatically accelerate sophisticated cyberattacks in sectors relying on complex, interconnected—and often decades-old—technology systems.
This regulatory moment reflects broader anxieties about technological dominance and national vulnerability. The government's willingness to delay a major commercial product release shows policymakers take these risks seriously, even as they ultimately approved the launch.
The Competitive Landscape
OpenAI's release comes as other AI companies navigate similar pressures. Anthropic abruptly disabled its most advanced models, Mythos 5 and Fable 5, for all users less than one month ago following the U.S. government's June 12 export control order over national security concerns. The restrictions were lifted last week after Anthropic implemented certain safeguards, ending a period of regulatory uncertainty that limited availability for users worldwide.
Meanwhile, Elon Musk's SpaceXAI announced it was making its leading model Grok 4.5 available to the public. Zhipu, trading as Knowledge Atlas Technology JSC, also launched its GLM 5.2 model last month, which is free to download, fine-tune, and run on an enterprise's own servers—a model that sidesteps some regulatory constraints by allowing local deployment.
OpenAI previewed GPT-5.6 in late June, touting improved agentic capabilities in coding, biology, and cybersecurity. The company noted that GPT-5.6 Sol was competitive with Anthropic's Mythos Preview on the ExploitBench cybersecurity benchmark, suggesting the models are approaching functional parity even as regulatory frameworks remain unsettled.
The White House and the U.S. Department of Commerce did not respond to requests for comment outside regular business hours. OpenAI also did not immediately respond to requests for comment.
Why This Matters:
This approval represents a critical test case for how democratic governments can maintain oversight of powerful technologies while allowing commercial innovation. The fact that the Trump administration required pre-release testing and limited initial access demonstrates that even administrations skeptical of regulation recognize frontier AI poses genuine risks requiring institutional oversight. The temporary restrictions on Anthropic's models showed these aren't empty threats—companies that don't comply face real consequences. Yet the fundamental tension remains unresolved: concrete standards for releasing frontier models haven't been finalized, meaning this approval happened within a regulatory gray zone. As AI capabilities accelerate and global competition intensifies, the framework established here will shape whether public institutions can meaningfully govern transformative technologies, or whether market forces and national security concerns alone determine what gets built and deployed.