OpenAI secured U.S. government approval to launch its most advanced AI model globally after a month-long delay triggered by national security concerns over potential misuse by foreign adversaries. The IPO-bound startup will roll out GPT-5.6 Sol, Terra and Luna on Thursday following additional testing and meetings with government officials, according to Axios. The White House and the U.S. Department of Commerce didn't respond to requests for comment outside regular business hours.
The approval comes as Washington and Beijing race to develop cutting-edge AI systems that experts warn could dramatically accelerate sophisticated cyberattacks against critical infrastructure. OpenAI had limited GPT-5.6 access to a small group of vetted partners whose details were shared with authorities after the delay last month. The company previewed the models in late June and touted improved agentic capabilities in coding, biology and cybersecurity. At the time, OpenAI said GPT-5.6 Sol was competitive with Anthropic's Mythos Preview on the ExploitBench cybersecurity benchmark.
The Security Dimension
Washington has increased scrutiny of advanced AI model releases to identify potential threats on concerns the technology could be misused by the military or the intelligence establishment in China, Russia and other countries. The United States and China are in a race to develop cutting-edge AI models the likes of which could dramatically accelerate sophisticated cyberattacks in sectors relying on complex, interconnected and often decades-old technology systems.
OpenAI competitor Anthropic had abruptly disabled its most advanced AI models, Mythos 5 and Fable 5, for all users after the U.S. government's June 12 export control order over national security concerns. The curbs were lifted last week after Anthropic implemented certain safeguards. While Washington has lifted export controls for Anthropic's Fable model, Mythos, which is designed for cybersecurity professionals, is still only available to some "trusted" U.S. organizations. In China, authorities are worried about the potential for Mythos to exploit software vulnerabilities and that the U.S. might deploy the model against Beijing's interests. Anthropic has warned it was "probably impossible" to make any AI model fully robust against jailbreaks.
Chinese authorities have also held meetings with top tech firms about potentially restricting overseas access to China's most advanced AI models, including those yet to be released.
New Voice Technology
OpenAI on Wednesday launched GPT-Live, a new family of voice models capable of listening and speaking simultaneously in real time. The company said it will roll out two versions of GPT-Live — GPT-Live-1 and GPT-Live-1 mini — to users globally on Wednesday. In May, OpenAI introduced three audio models for its developer platform, aiming to make voice-based software agents more conversational and capable of completing tasks in real time.
Billionaire Elon Musk, whose SpaceXAI rivals Anthropic and OpenAI, said on Wednesday his company was also making its leading model Grok 4.5 available to the public. Musk's AI startup xAI was acquired by SpaceX in February. He said in May that xAI would cease to exist as a separate company and would instead become SpaceXAI.
Why This Matters:
The approval marks a shift in how Washington manages AI competitiveness and national security. European policymakers face the same dilemma: how to foster innovation while preventing adversaries from weaponising advanced models. The U.S. has chosen a case-by-case review system that allows companies to launch after demonstrating safeguards. Europe's AI Act, by contrast, imposes blanket regulatory requirements that risk slowing deployment and handing competitive advantage to American and Chinese firms. The cybersecurity capabilities of models like GPT-5.6 and Anthropic's Mythos aren't theoretical — they can identify vulnerabilities in critical infrastructure at scale. Europe must balance open research with export controls on the most sensitive capabilities, or risk falling behind in both AI development and defence readiness.