Google says it has identified and stopped the first known zero-day exploit developed with the help of artificial intelligence. The finding comes from the Google Threat Intelligence Group, which flagged the vulnerability before it could be used in what researchers described as a planned mass exploitation event targeting an unnamed open-source, web-based system administration tool.
The attackers, described as prominent cybercrime actors, were aiming to use the exploit to bypass two-factor authentication. Researchers spotted telltale signs of AI involvement in the Python script behind the attack. Among them: a hallucinated CVSS score and unusually structured, textbook-style code formatting consistent with output from a large language model.
The discovery marks a notable shift in the threat landscape. Security researchers have long warned that AI would eventually lower the bar for sophisticated attacks. This appears to be the first confirmed case where that warning became reality. Google has not named the target system or the threat actors involved.
