A successful detection result would indicate that Claude may have processed content. It would not prove that Claude was the original author, because users can submit human-created text, research, images, or files to Claude for summarizing, editing, translation, or proofreading.
Likewise, the absence of a watermark should not be treated as proof that a human wrote content. Older Claude models may not yet support marking, short text may contain too little material for reliable detection, and heavily edited output may lose its signal.
Not exactly bulletproof, is it?