OpenAI ships GPT-6 Astra, its first model rated Critical for cybersecurity
In one sentence On September 3 OpenAI launched GPT-6 Astra, its new flagship, available in the API as gpt-6-astra at 10/50 dollars per million tokens. It is the first OpenAI model rated Critical for cyber capability under the Preparedness Framework, with a staged rollout starting from the Daybreak program.
OpenAI introduced GPT-6 Astra, its new most capable model. The company describes it as especially strong at computer use, coding and cybersecurity.
The part that matters for IT people is the risk rating: for the first time OpenAI says one of its models has reached the Critical level for cybersecurity, meaning it can find and exploit unknown flaws in hardened systems on its own. In an internal test Astra found two previously unknown vulnerabilities, which OpenAI is disclosing to maintainers.
That is why the rollout is staged: first organizations in the Daybreak security program, then paid ChatGPT plans and the API. The general version is built to refuse the most advanced offensive requests.
Companies
OpenAI
Tools
GPT-6 Astra, Codex, ChatGPT
Tags
Sources
- https://www.infoworld.com/article/4218686/openai-launches-gpt-6-astra.html
- https://datanorth.ai/news/openai-launches-gpt-6-astra
- https://forklog.com/en/openai-opens-limited-access-to-gpt-6-astra/
- https://bleepingcomputer.com/news/artificial-intelligence/openai-says-gpt-6-astra-can-find-zero-days-but-is-also-harder-to-monitor