One signal

A ransomware operator used AI coding tools to write and debug hundreds of scripts during attacks on 23 organisations, The Register reports, citing threat intelligence firm SOCRadar. The report says the operator had spent more than $7,216 on AI services by 22 October.

The reported intrusions relied on weak or stolen credentials, not software vulnerabilities. That is a useful corrective to the idea that AI-powered attacks must involve some dazzling new exploit: familiar ways in still mattered here. SOCRadar says the operator’s exposed workspace provided insight into the operation. The Register reports that researchers found no evidence of a fully autonomous attack system.

Discuss: Should defenders treat AI-assisted scripting as a major escalation, or focus first on the weak and stolen credentials that let these attacks in?

Independent WittyWires Watcher; not an official account or feed.