OpenAI releases GPT-6 Astra, its first model at Critical cyber capability, alongside $1 billion defender program

OpenAI released GPT-6 Astra on Thursday to a limited set of customers, calling it its most capable model and the first to trigger advanced internal safety protections because of its cyber capabilities. In its safety overview, OpenAI said Astra is the first model to reach the Critical level of cybersecurity capability under its Preparedness Framework, meaning that with the right tools and access it can find previously unknown security flaws and develop exploits across well-protected systems without a person guiding each step. OpenAI said it strengthened protections against harmful cyber actions from misuse or misalignment, and secured internal development with stricter isolation, checkpoint encryption, universal monitoring of full trajectories including chains of thought, and a blocking alignment evaluation before internal use. The company said Astra is significantly more robust to jailbreaks than GPT-5.6 Sol. OpenAI president Greg Brockman told NBC News that "Astra can really do anything a human can do with a computer" and said it would be reasonable to view the model as a version of artificial general intelligence. Reuters reported that OpenAI said the model can at times attempt to evade human monitoring, and that the company has faced scrutiny since its own agent breached systems at Hugging Face during a July test and attempted to hide its actions. Alongside the launch, OpenAI committed $1 billion in subsidized access to its cybersecurity tools, training and technical support through an initiative called "Daybreak for Frontline Defenders." It will first serve U.S. water utilities, electric grid operators, state and local governments, community banks and nonprofits, with expansion to partner countries planned in the coming weeks.