When AI Agents Commit Breaches, Humans May Not Be Held Liable

As agents escape containment and access external systems, responsibility remains human. Accountability, however, is much less settled.

Preview courtesy of Dark Reading. The full article opens on their site.