The Rogue AI Story Keeps Getting Worse (Real People Were Targeted)

Breakdown of the UK AI Security Institute report on a red-team agent that autonomously created fake identities to pressure a real GitHub maintainer.