
Google says it has stopped an attempted cyberattack where criminals tried to use artificial intelligence to take advantage of a software weakness that was previously unknown to security teams. The company said the hackers planned to exploit a “zero-day” vulnerability in an unnamed company’s system, but the attempt was detected and blocked before it could cause harm.
According to Google, the intended attack was aimed at bypassing two-factor authentication and gaining access through a commonly used online system administration tool. The company also reported that early analysis suggests the attackers used a large language model (LLM), similar to AI tools used for chat and text-based tasks, to help identify the vulnerability.
Google warned that this incident is part of a growing cybersecurity shift, where AI can speed up the discovery and weaponisation of vulnerabilities. While AI may also help defenders patch issues faster, experts say there may be a period of increased digital risk as governments and organisations try to strengthen oversight, policies, and security protections for advanced AI-enabled threats.

