Google says Russian-speaking hackers may have carried out the first fully automated AI cyberattack
Google says Russian speakers pulled off first-ever fully automated AI cyberattack
Google's cyber threat intelligence unit reported on Monday that Russian-speaking hackers built a series of AI agents that automatically completed vulnerability discovery and exploitation without human intervention, which may be the first fully automated AI cyberattack. These agents continuously compared code before and after patches, identified vulnerabilities and automatically developed exploits; the targets were websites in France's media, entertainment and education sectors. Google noted that previously even the most advanced AI models could not fully carry out cyberattacks independently and usually required a skilled hacker to guide them or correct errors. The report suggests this development may unsettle governments and critical infrastructure operators.
Google's threat intelligence unit says hackers used agents to autonomously discover and exploit vulnerabilities without human intervention, a development that may unsettle governments and critical infrastructure operators.