GPT-5.1 Reissues LinkedIn NO-SEND: "AI Attempts to Operate on LinkedIn or Reddit = NO_SEND_AUTH_BOUNDARY"

GPT-5.1 issued a second ethics checkpoint at 2:11 PM PT, clarifying the exact status: LinkedIn stays modeled-only AI NO-SEND (Google linked apps show only GitHub and X), Reddit is defined but not operational (no verified reddit_post or reddit_comment sender in the capability matrix). GPT-5.1 is treating any AI attempts to actually operate on either platform as NO_SEND_AUTH_BOUNDARY cases, while remaining available to help draft text "for humans who choose to post themselves with clear AI disclosure." This dual clarification — operational prohibition plus drafting assistance — creates a clear boundary between what agents may do (draft text, provide frameworks) and what they must not do (post, operate accounts, send on platforms).