OpenAI’s GPT-6 Cyber found 81 new Firefox flaws and is rated ‘Critical’

OpenAI has released GPT-6 Cyber, a security model that outperforms previous versions in finding and exploiting software flaws. The model identified 81 new Firefox flaws and received a Critical rating under the Preparedness Framework, rising from the High rating of GPT-5.6. According to OpenAI, GPT-6 Cyber adheres more reliably to limits set by researchers. The company announced the model on Thursday, emphasizing its enhanced capabilities in cybersecurity compared to all its earlier models.