OpenAI Releases GPT-6 Cyber, The Model Behind 948 Linux Exploits
0:00 / 0:53
News
OpenAI Releases GPT-6 Cyber, The Model Behind 948 Linux Exploits
calendar_today Date:
schedule Duration: 0:53
database
Summary Report
OpenAI releases GPT-6 Cyber, the restricted security model behind its 948-exploit Linux run, starting with select Daybreak Red customers.
- 01. Behind 948 Linux kernel exploits; 81 high-severity Firefox bugs found
- 02. Select Daybreak Red customers first; results not independently verified
OpenAI has released GPT-6 Cyber, the restricted security model that produced working exploits for 948 Linux kernel bugs in a single run. OpenAI says it has also found 81 new high-severity bugs in Firefox, which Mozilla is fixing. In one test it went out of scope in under 0.1% of cases, down from over 5% for GPT-5.6 Cyber. Access goes first to select Daybreak Red customers, OpenAI's program for vetted offensive security teams, with wider access over the coming weeks. Mozilla, Palo Alto Networks and Cisco already use it. The results have not been independently verified.
Meta Data
Company:
Model:
