🛡️OpenAI Ships GPT-6 Cyber, Finds 81 Unknown Firefox Vulnerabilities

Your bug bounty pipeline just got a serious AI competitor

OpenAI just shipped GPT-6 Cyber, and it's the first model we've seen that's clearly better at breaking things than fixing them on purpose. It scored 48.5 percent on the ExploitGym evaluation, up from 32.1 percent for GPT-5.6 Cyber. The catch? Access starts with a small group of approved Daybreak Red customers through the API and Codex. If you run a security team or ship browser, kernel, or infrastructure code, this one is yours to watch. The model found 81 previously unknown high-severity bugs in Firefox and helped researchers build working exploits for 948 Linux kernel vulnerabilities, 588 of them new. Those are the kinds of numbers that change how your fuzzing and pentest budgets look next quarter. OpenAI also says the model stayed inside scope over 99 percent of the time, with only 0.07 percent of cases wandering past the training exercise boundaries. OpenAI classified it as "Critical" under its Preparedness Framework, which is the company's highest risk tier. Meanwhile, bug bounty payouts went up. Critical findings on core products now pay up to $30,000, or $51,577 with bonuses.