OpenAI Unlocks ChatGPT 5.6 After White House Cybersecurity Review

OpenAI launched its newest AI model, ChatGPT 5.6, on Thursday, ending a delay triggered by cybersecurity concerns from the Trump administration. The White House had asked the company last month to limit access to a small group of government-approved users, and OpenAI complied, restricting the model to trusted partners while briefing officials on its capabilities. The wider release followed additional testing by the government's Center for AI Standards and Innovation agency.

The staggered rollout mirrors similar restrictions placed on rival Anthropic's latest models, which faced a temporary export ban last month. Both cases have highlighted rising cybersecurity risks tied to advanced AI, strained relationships with U.S. allies over shared technology, and criticism of the government's ad hoc approach to regulation. ChatGPT 5.6, which includes a flagship product called Sol, is being billed as OpenAI's safest and most capable model yet, directly competing with Anthropic's Claude Fable and Mythos models that hit the public market earlier this month.

The Trump administration has largely pushed for rapid AI development, citing a need to outpace China, but recent fears of AI-driven cyberattacks have prompted more intervention. An executive order last month created a voluntary review process for AI firms to submit models for evaluation. The restrictions on who can access these tools have unsettled industries that rely on them, adversarial governments facing new threats, and allies excluded from early access. OpenAI initially limited ChatGPT 5.6 to U.S. entities. Both OpenAI and Anthropic are racing toward initial public offerings later this year, with Anthropic valued at $965 billion after a May funding round and OpenAI reportedly eyeing a $1 trillion valuation, though its IPO may slip to next year amid tech market volatility.

Source: The Guardian