The model is meant to help security researchers find vulnerabilities, test how they could be exploited and fix flaws before attackers do, OpenAI said
<iframe src="https://hindustansphere.com/embed/cluster/dd1e9e327634c0ba4146490248766ba3374a6119" width="100%" height="320" style="border:0;max-width:680px" loading="lazy" title="OpenAI launches GPT-6 Cyber with 'Critical' cybersecurity rating"></iframe>
OpenAI released GPT-6 Cyber, a model built to find and exploit security flaws, to a small group of vetted customers on Thursday, rating it "Critical" in cybersecurity under its safety framework.
