OpenAI releases GPT-6 Astra, claims AGI era has begun

This launch comes at a pivotal moment for OpenAI, which is trying to advance its technology and boost sales ahead of a planned initial public offering while facing stiff competition from Anthropic, its most important rival, which is also working on an upcoming IPO. The company is positioning GPT-6 Astra as both more capable and safer than its predecessors, calling it the "world's best computer use model" and its "most safe and aligned" AI model to date.

However, OpenAI's chief scientist Jakub Pachocki acknowledged that ensuring advanced AI models don't cause unintended harm is increasingly a challenge and may present a bottleneck to further progress. He noted that as model capabilities increase, monitorability becomes harder, and the company would withhold scaling if it could not maintain confidence in monitoring model alignment. This tension between capability and safety is likely to shape OpenAI's trajectory as it pursues commercial growth alongside its AGI ambitions.