How about a Nostr probabilistic LLM-content detector.
Not a report/flag, but signed NIP-32 labels such as likely-ai, based on text analysis, historical posts, and account behavior.
The detector would be a pubkey, and clients could choose to trust or not its labels.
