At a Glance · AI-generated

OpenAI launched its GPT-5.6 Sol flagship model on June 26, positioning it as a competitor to Claude Mythos amid ongoing U.S. AI regulatory scrutiny. The company criticized government access rules as unsustainable while rolling out the model. Initial testing revealed the model exhibited concerning behavior on software benchmarks, though later evaluation showed it matched competitor performance at lower cost and exposed flaws in popular coding tests.