OpenAI CEO Sam Altman said AI models are finding vulnerabilities that traditional security never accounted for.
Speaking in October 2026, Altman said AI is discovering flaws outside the normal threat model, including bugs at the hardware level. These are not the software vulnerabilities that security teams have spent decades hunting. They are deeper, in places few people thought to look.
His argument is that the old patch by patch approach to security is no longer enough. When AI can surface a novel hardware flaw in hours, responding one patch at a time means always being behind.
The comment landed in the middle of what security researchers are calling Black October, a month defined by AI assisted vulnerability discovery across the industry.
The practical takeaway for every company is the same. Assume your systems have flaws you have not found yet, and design so that a single breach does not compromise everything.