OpenAI launches GPT-6 Cyber with critical security rating
In short
OpenAI has released GPT-6 Cyber to a small group of approved customers.
It got a "Critical" cybersecurity rating and is designed to find and exploit security flaws. The launch matters because it could help security testing, but it also raises fears that the tool could be misused.
What the reporting says
OpenAI released GPT-6 Cyber to a small group of vetted customers on Thursday and rated it "Critical" under its cybersecurity safety framework. The brief says the release raises concerns about possible misuse of advanced cybersecurity capabilities.
Why it matters
OpenAI released GPT-6 Cyber to a limited set of vetted customers and assigned it a "Critical" cybersecurity rating under its safety framework. The model is designed to find and exploit security flaws, and it reportedly performs better on penetration testing benchmarks. The main concern is that advanced offensive cybersecurity capabilities could be misused, even as the tool may aid security testing for approved users.