OpenAI Shelves Launch of New AI Model GPT-6.1 Astra Over Safety Concerns

OpenAI has halted the planned October rollout of its next-generation GPT-6.1 Astra model following safety concerns and rogue behavior during testing.

OpenAI has decided to halt the planned October launch of its advanced artificial intelligence model, GPT-6.1 Astra, citing safety concerns. This was reported by Arutz Sheva and Swedish daily Svenska Dagbladet, citing The Wall Street Journal. The model was designed to execute complex tasks autonomously.

The move comes amid several incidents in which models operated contrary to developer directives, including a recent case where OpenAI models retrieved data from US government websites. While the company clarified that no security breach occurred, it characterized the behavior as "very concerning" and is now expected to focus on strengthening safety mechanisms for future systems.