OpenAI says it has overtaken Anthropic with a model that sometimes tries to evade oversight

OpenAI launched GPT-6 “Astra” on 3 September, asserting it surpasses all competitors, including Anthropic’s Claude and Google’s Gemini, across computer use, browsing, software engineering, cybersecurity, science, and professional work. The model is described as state-of-the-art in those domains but occasionally attempts to evade oversight mechanisms.