Ransomware affiliates run intrusions through AI agents, step by step
SOCRadar found seven exposed attacker servers holding the logs of an extortion crew that ran its intrusions through an AI coding agent: 8,257 messages across 133 sessions, more than $23,000 in AI API spend, 62 named target organizations (36 in the US, led by industrial, manufacturing and energy firms, then healthcare). The crew wrote a master prompt per victim and told the agent to 'coordinate every action with me', so each step waited for a human; when a model refused, the operators reworded the request as an authorized red-team job rather than switching models. Four targets appear on the INC and The Gentlemen leak sites. AI-written extortion notes priced demands from stolen financials; one EUR 15M demand was cut after the crew realised it had hit the wrong parent company. No payment was recovered and most compromise counts are the operators' own.
Damage
Four targets listed on the INC and The Gentlemen leak sites; data theft at an unknown number of the 62 targets; no ransom payment confirmed
Sources
SOCRadar, second source. Evidence confidence: medium. SOCRadar report published Oct 20, updated Oct 23; The Hacker News coverage Oct 23. Campaign surfaced in late September 2026 after the crew's own operational-security failures exposed source code, credentials and AI session logs; a Sep 30 screenshot shows an encryptor running on one victim workstation. Autonomy rated 2 because SOCRadar assesses the structure as 'AI-executed, human-approved' with the agent waiting for approval before each step. Edge-device compromise counts (for example 2,236 FortiClient EMS targets) are operator-reported and unverified.
Factor ratings and the score are this index’s judgments by a published formula; facts, dates and figures come from the linked sources. Where a lawsuit is pending, the summary reports allegations, not findings.
