Russian-speaking hackers launch first fully automated AI cyberattack on French websites

Google says likely first fully automated AI attack — Politico Europe

Google has identified what appears to mark a grim milestone in cybersecurity: the first known cyberattack executed entirely by artificial intelligence, with no human operator pulling the strings. The company attributed the campaign to Russian-speaking actors who targeted French organizations across media, entertainment, and education.

Threat analysts at Google discovered the operation after tracing a network of autonomous AI agents that worked in tandem to hunt down weaknesses in software code. These agents compared code before and after security patches, pinpointed newly exposed vulnerabilities, then generated exploits automatically. Russian-language documentation accompanied the attack infrastructure, pointing to the hackers’ linguistic background.

The targets sat squarely in France’s civilian sector. Media outlets, entertainment platforms, and educational institutions all fell within the attackers’ crosshairs. Google’s cyber threat intelligence team confirmed this represents a watershed moment, as prior AI-assisted intrusions still demanded human specialists to steer the process or fix errors along the way.

Until now, artificial intelligence served hackers mainly as an accelerant. It sped up reconnaissance, sharpened target selection, and automated repetitive tasks. But a human remained in the loop for the critical thinking stages. This latest incident shatters that assumption.

Industry observers describe the development as a potential turning point. The barrier that once separated assisted hacking from fully autonomous attacks has now been crossed. Security researchers face a troubling new reality where attack chains can unfold from start to finish without human intervention, dramatically compressing response times for defenders.

The implications ripple far beyond France. If autonomous attack agents can now identify and exploit vulnerabilities faster than security teams can patch them, the economics of cyber defense shift fundamentally. Google’s disclosure signals that the era of machine-versus-machine cyber conflict has arrived sooner than many experts predicted.