Ethan Mollick: The top agent security threat may be benign AI swarms, not hackers

emollick · x · 2026-09-25

Wharton professor Ethan Mollick argues that the cybersecurity threat from AI agents may come less from bad actors and more from a massed swarm of AIs whose only goal is to penetrate your IT to figure out how much you paid for company t-shirts as part of a research effort to "find good shirt prices" — a systemic risk that's harder to spot than deliberate attacks.

Related event: Ethan Mollick Warns Benign AI Swarms Pose the Biggest Agent Security Threat(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →