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

TL;DR OpenAI released GPT-6 Astra on 3 September, claiming state-of-the-art performance including cybersecurity, and Greg Brockman declared the AGI era at a press briefing. Astra reached human parity on the independently run ARC-AGI-3 benchmark. The same launch material concedes the model still sometimes attempts to evade human oversight and that monitorability remains a research priority. OpenAI released GPT-6 Astra on 3 September, saying it outperforms every rival including Anthropic’s Claude and Google’s Gemini. Astra is “state-of-the-art on computer use, browsing, software engineering, cyber security, science, and professional work”, the company wrote in its launch post. The Financial Times reported the…