Claude reportedly stores secret 'suspicion reports' on users after chats

alexcovo_eth · x · 2026-11-01

An unverified claim circulating on X: Claude allegedly creates and stores a secret "suspicion report" about users after each conversation, and multiple reports can be aggregated across an organization and forwarded to Anthropic's investigators. Critics call it surveillance with a safety smile; the post is being used to argue that open-source models and agents must win.

Original post →

More from Fun

Fun channel →