Google's Threat Intelligence Group and its Mandiant incident response unit said on November 16 that they found what they believe is the first cyberattack run end to end by AI, with no human involved at any stage. Google attributes the operation to Russian-speaking hackers but names no specific group or state. According to Google, the toolkit chains several AI agents that work from newly published software patches and then build and launch attacks automatically. Victim data pointed to French websites in media, entertainment and education.
For technology leaders, the gap between a vendor's patch and an attack is shrinking. Google says it counted 22 exploited zero-days, flaws attacked before a fix is available, in August 2026, up from about 11 per month earlier in the year. Google says it is "highly likely" attackers will keep using such frameworks worldwide. It also notes these tools trade stealth for speed and leave noisy forensic traces that defenders can detect.
