Google Threat Intelligence Group has reported the first autonomous cyberattack using artificial intelligence to exploit vulnerabilities in French media, entertainment, and educational websites. The attack marks a significant development in AI-enabled cybercrime.

The cyberattack employed a framework known as 'exploit_agent,' which autonomously identified and exploited software flaws.

Evidence found on the exposed server included Russian-language instructions, suggesting a possible origin of the attackers. The tools discovered on the server were also capable of supporting further stages of cyberattacks, such as gaining network access and removing rival web shells.

The incident has prompted a coordinated response from the European Union, which recently sanctioned individuals linked to Russian cyber operations.

Google has noted a rise in vulnerability disclosures and zero-day exploits in 2026, indicating an increasing trend in cyber threats.