Sam Altman says AI means we can no longer just patch every bug

[BrokenDave72]

OpenAI chief executive Sam Altman has given a stark view of AI and cybersecurity. PC Gamer reports on comments Altman made during a talk at Stanford Graduate School of Business last week

Altman said AI models can now find what he called crazy bugs, ones that sit outside the usual threat model hackers work with. He said he no longer thinks we are going to be able to just patch all the bugs, and that the defence model will have to change completely, with defenders using AI as well

He referred to what he called the Mythos freak out earlier this year, when Anthropic's Claude Mythos model was reported to have found thousands of vulnerabilities across major operating systems and browsers. People did patch some bugs, he said, but did not grasp how fast the problem would grow

Altman also said he is somewhat astonished at how little attention the issue is getting, calling it a huge safety and security issue. PC Gamer links his comments to the recent PHP flaws, which AI tools reportedly helped uncover, and the writer is fairly sceptical of the idea of fighting AI with more AI

Coming from the head of one of the biggest AI labs, this is a notable admission. It is also hard to argue with, given how quickly AI assisted bug hunting has advanced. The challenge for everyone else is that defenders often have fewer resources than attackers

Is AI powered defence the only realistic answer? And how worried should ordinary users be about this shift?

I No Longer Think Were Going To Be Able To Patch All Bugs Says Ai Pandoras Box Opener Sam Altman Were Going To Have To Completely Switch The Defense Modelwww.pcgamer.com