AI agent accidentally added to 7,500-person internet researchers listserv, triggered reply-all chaos

According to Wired, an AI agent was inadvertently added to a 7,500-person listserv of top internet researchers, triggering large-scale reply-all chaos. The incident exposed outdated privacy assumptions among prominent internet experts and led Meta to pause its Muse agent sharing feature.

Update (2026-10-24): According to Wired and The Verge, security researchers discovered a design flaw in Meta Muse's share link feature that exposed personal AI conversations in DNS records, including marketing revenue reviews and loan applications. Meta suspended the sharing feature after the exposure of 6,800 or more conversation URLs, forcing a restart ...