OpenAI unveils GPT-6 Cyber security model
Original title: OpenAI Says New GPT-6 Cyber Model Hunts Flaws Without Crossing the Line
GPT-6 Cyber achieves 48.5 percent on ExploitGym and stays within scope in 99.93 percent of tests.
OpenAI introduced GPT-6 Cyber, a model for security research that finds more vulnerabilities in Linux and Firefox. It reproduced 1,724 Linux kernel vulnerabilities and developed exploits for 948 of them. The model avoids unauthorized actions in alignment tests. Access is limited to Daybreak Red customers.
- 48.5 percent on ExploitGym
- 1,724 Linux kernel vulnerabilities reproduced
- 81 new Firefox flaws documented
- 99.93 percent alignment rate
- Vendors
- OpenAI
- AI models
- GPT-6 Cyber
From the original article
2 min read at securityweek.com
Rated Critical for cybersecurity, OpenAI’s new model beats GPT-5. 6 Cyber at exploitation while staying in scope more reliably. The post OpenAI Says New GPT-6 Cyber Model Hunts Flaws Without Crossing the Line appeared first on SecurityWeek.
By Eduard Kovacs · securityweek.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.