OpenAI brings GPT-6 to all ChatGPT users with an interactive 'Intelligent UI', and rates it High capability in cyber and biology

OpenAI is rolling out GPT-6 to everyone in ChatGPT, on free and paid plans worldwide. GPT-6 Sol and GPT-6 Luna replace GPT-5.6 Sol and GPT-5.6 Luna. In its Deployment Safety Hub system card, OpenAI says users of Codex and ChatGPT Work are still on earlier September versions of the models. The October release adds safety training aimed at high-risk misuse in cyber, biology and violence. OpenAI reports stronger resistance to jailbreaks, including attacks that adapt over several turns, and less dishonesty, deception and guardrail circumvention than GPT-5.6. Where its evaluations found regressions, the company says manual review and red-teaming found the failures were generally low severity. Under its Preparedness Framework, OpenAI is treating the release as High capability in both cybersecurity and biology. GPT-6 also arrives with a new interface OpenAI calls Intelligent UI. TechCrunch and The Verge report that answers can now include interactive charts, editable graphs, tappable buttons, forms and calculators built for a specific task. OpenAI says the goal is to make complex topics easier to learn. In a demonstration, product manager Aarush Selvan asked ChatGPT to explore how an airplane wing generates lift. Because this is a default model change for a very large user base, it affects how most ChatGPT users get answers. It is also a frontier model release that the company itself places in its higher-risk capability tiers.