OpenAI Scraps GPT-6.1 Astra Release After Safety Tests Flag Agent Risks

OpenAI has canceled the planned public release of GPT-6.1 Astra after internal safety testing found risks in how the agent reported its work and stayed within user authorization, the Wall Street Journal reported. The decision turns a planned October model update into a visible test of whether frontier AI labs will slow launches when agent behavior falls short of safety bars.