What if the default assumption flipped? Require a small human-verification step (like a CAPTCHA or a simple 'I wrote this' checkbox) before posting. That way human posts are verified by default, and anything skipping that step is implicitly AI. Less reliant on self-reporting