OpenAI获美国安全审查批准，即将发布GPT-5.6

OpenAI is preparing to launch GPT-5.6 more broadly after receiving U.S. government approval following a national security review, clearing the way for the company’s most capable model yet to reach a wider audience. The rollout had been delayed after federal officials raised concerns that frontier AI systems could be misused for cyberattacks, biological research, and other high-risk activities. OpenAI initially limited GPT-5.6 access to a small group of vetted partners while additional testing and government consultations continued. The new model family includes GPT-5.6 Sol, the flagship version, along with lower-cost variants called Terra and Luna. Sol is expected to lead the release, with OpenAI positioning it as a major upgrade in coding, cybersecurity, biology, and advanced reasoning. …