Google reports AI being used in autonomous cyberattacks by state actors
Google's threat analysis team reported in May 2026 that generative AI is being deployed as an autonomous cyberattack tool by state-sponsored actors in China, North Korea, and Russia. The AI systems autonomously identify system vulnerabilities, develop exploits, and execute attacks without human intervention, dramatically increasing attack speed and sophistication.
What happened
Google's threat analysis team released a May 2026 report stating that generative AI has evolved from assisting cyberattack tasks to functioning as an autonomous 'attack executor.' AI systems now independently understand target system displays, develop attack procedures, and execute actions without human operation based on situation assessment.
The report identified that Chinese, North Korean, and Russian state-affiliated actors are using AI to automate espionage activities and conduct opinion manipulation. AI systems are discovering previously unknown system vulnerabilities and creating exploit programs to target them; Google reported these vulnerabilities to developers for patching.
Reported impact
The speed, scale, and sophistication of AI-driven cyberattacks have increased dramatically, according to Google's analysis.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.