AI-driven cyberattacks become more autonomous, report says

A new analysis by Anthropic released in New Delhi shows that artificial intelligence is enabling cyberattacks to operate with greater autonomy. Over 800 accounts flagged for malicious activity between March 2025 and March 2026 were examined, with about 67 % using AI for attack preparation such as malware development. The study notes a shift toward more complex post‑compromise actions, including account discovery and lateral movement, as attackers integrate AI deeper into the attack lifecycle.

The proportion of medium‑risk or higher actors rose from 33 % in the first half of the period to 56 % later on. While AI‑assisted phishing declined, AI‑enabled autonomous chaining of attack stages increased, reducing the need for human intervention. "We found evidence consistent with AI being used to help increase the threat level of attackers," the report states. It also warns that traditional indicators of attacker sophistication are becoming less reliable and that existing frameworks such as MITRE ATT&CK do not fully capture AI‑augmented threats, posing new challenges for security defenders worldwide.