Google thwarts AI attack bypassing 2-factor authentication on admin tool

Technology May 11, 2026

Google just stopped a cyberattack where hackers used AI to find a hidden security flaw and break past two-factor authentication on a popular admin tool.
No harm was done, and Google quickly let the affected company know what happened.

Google warns AI eases vulnerability discovery

The attackers used a large language model (Google said it was most likely not Google's Gemini or Anthropic's Claude Mythos) to spot the vulnerability.
Google said there was no evidence it was tied to an adversarial government, countries like China and North Korea are exploring similar tactics.
John Hultquist from Google called it "It's here," warning that AI makes it much easier and faster to find and exploit weaknesses.
This is a wake-up call for stronger digital defenses.