According to @emollick, no watermark can be foolproof against advanced AI, highlighting technical limits and policy needs, as reported by Twitter.

According to @emollick, subjective recognition of AI prose clashes with unreliable detection tools, complicating proof of AI use.