AI found about a thousand vulnerabilities in Chrome — Google is racing to patch them

ER10 reported a telling story: AI-based automated vulnerability discovery tools found on the order of a thousand security issues in the Chrome browser, and Google is shipping patches at an accelerated pace. Some findings are minor, but the scale speaks for itself: algorithms comb through a codebase in a way no team of human auditors could resource.

The coin has two sides. Attackers use the same technology: an AI that finds a vulnerability for the defender finds it for the intruder too — the only question is who reaches a given hole first. The race between "find and patch" and "find and exploit" has sped up severalfold, and the lifetime of an unpatched flaw is shrinking from months to days.

For an ordinary company in Kazakhstan this yields a simple rule that closes the lion's share of the risk: updates must install automatically, everywhere — browsers, operating systems, routers, the website CMS. The incident of "we got hacked through a hole patched six months ago" is the most galling and the most common scenario. Half an hour spent checking your auto-update policy costs less than any incident post-mortem.