The comment I replied to said this:

> As AI bans go, this one doesn't seem too stringent if you disclose AI and actually check what it's doing.

I'm not arguing for or against the policy, but the policy as-written reads to me a complete and total ban on the submission of any and all generated text, which is not what that commenter suggested. I'm not sure you could possibly make it more stringent without getting into the territory of thought police.