AI Is Helping Low-Skill Hackers Pull Off Advanced Cyberattacks
Help Net Security, Friday, June 5th, 2026
Anthropic's analysis finds AI enabling less-skilled threat actors to run sophisticated cyberattacks.
Anthropic examined 832 banned accounts engaged in malicious cyber activity between March 2025 and March 2026, mapping their behavior to the MITRE ATT&CK framework.
The research found 67.3% of reviewed accounts used AI for preparatory tasks like malware development, with post-compromise techniques once restricted to skilled actors now accessible to less sophisticated operators.
During the study period, AI-assisted account discovery rose 8.9% while phishing assistance declined 8.6%. Medium- and high-risk actors using AI for lateral movement and credential dumping increased from 33% to 56% between the study's two halves.
Researchers concluded that emerging agentic systems and autonomous killchain orchestration represent gaps in current threat intelligence frameworks requiring an updated taxonomy.