Signal president Meredith Whittaker warned that the rise of AI agents — systems that act autonomously and access broad swaths of user data — poses an "existential" threat to secure messaging apps and the privacy guarantees they provide. Whittaker argues that agents require access to sensitive information (contacts, messages, credentials), creating new attack surfaces and undermining app-level end‑to‑end encryption. Her comments stress architectural tradeoffs as major operating‑system vendors and platform companies integrate agents. Campus IT and security teams should evaluate how agent frameworks change threat models for university messaging, research data and privileged communications, and update access controls, vendor assessments and incident‑response playbooks accordingly. (Source: interview with Signal’s president at a technology conference.)