OpenAI releases GPT-6 Astra, its first model to hit a Critical cyber threshold

OpenAI released GPT-6 Astra on Thursday to a limited set of customers, calling it the most capable model it has ever broadly deployed. According to OpenAI's system card, Astra is the first model to reach the Critical level of cybersecurity capability under the company's Preparedness Framework, meaning it can find previously unknown security flaws and develop exploits across well-protected systems without a person guiding each step. OpenAI says it responded with stricter isolation, checkpoint encryption, universal monitoring of full trajectories including chains of thought, and a blocking alignment evaluation before internal use. The company also says Astra is significantly more robust to jailbreaks than GPT-5.6 Sol. The Hacker News reported that Astra scored 100 percent on ExploitBench and that OpenAI is blocking requests for proof-of-concept exploits. TechCrunch reported the model went first to customers in OpenAI's Daybreak cybersecurity program, with Pro, Plus, Enterprise, Business and API access to follow over the next week. OpenAI president Greg Brockman told NBC News the model can do anything a human can do with a computer and said it would be reasonable to view it as a version of artificial general intelligence. 9to5Mac described the release as a major upgrade to ChatGPT and Codex. Within hours, The Verge reported that CEO Sam Altman apologized for a messy rollout after paying users expecting access were left waiting, with no clear timeline offered.