← Thursday, August 13, 2026
AI Frontier · Aug 11
ModerateOpenAI builds an AI that hunts for security holes
OpenAI made a new AI called GPT-5.6-Cyber. It is built to find weak spots in software before bad actors do. In tests it found two unknown bugs in Google's Chrome browser, which are now fixed. OpenAI only gives it to approved security teams, not the public. It is a sign AI is getting good at both breaking and guarding computer systems.
Why it matters to GCs
Builders get hit by hackers too, and a stolen bid or a fake payment can cost a lot. Tools like this help defenders find and fix holes faster. It is also a warning to keep your firm's logins and data locked down.
SecurityWeek·Via POLR AI Hub