OpenAI releases GPT-6 Cyber for authorized vulnerability research

OpenAI says its security-focused model found 81 new high-severity Firefox vulnerabilities and developed working exploits for 948 Linux kernel bugs. The company also says the model follows researchers’ boundaries more reliably than its predecessor. According to SiliconANGLE, GPT-6 Cyber is a version of GPT-6 Astra trained specifically for authorized vulnerability research.

• RuntimeWire reports that OpenAI says the gated model scored 48.5% on ExploitGym.

Covered by 5 publishers within 19 hours of the first report.

• RuntimeWire: OpenAI launches GPT-6 Cyber for gated offensive security research

Reporting5

RTW RuntimeWire OpenAI launches GPT-6 Cyber for gated offensive security research · Ryan Merket 12:12 PM

SA SiliconANGLE OpenAI launches GPT-6 Cyber, an Astra model built to keep vulnerability research in scope · Duncan Riley 5:52 PM

CSN Cyber Security News OpenAI Launches GPT-6 Cyber That Finds 81 Firefox Flaws and Develops 948 Linux Exploits · Guru Baran 9:04 PM

AH Android Headlines OpenAI Unveils GPT-6 Cyber to Find Vulnerabilities Before Hackers Do · Jean Leon 9:15 PM

SAF Security Affairs OpenAI Rates GPT-6 Cyber Critical After Major Vulnerability Discoveries · Pierluigi Paganini 7:26 AM

Related

CS OpenAI launches GPT-6 Cyber with tighter controls on security testing November 5

OAI Franklin Templeton develops OpenAI workflows for equity research November 4

WCN OpenAI Tests GPT-6.2 Astra Amid GPT-6 Robotics Dominance October 31

UAI OpenAI Makes Decisions API Generally Available with GPT-6 Luna October 22

Topics OpenAI