Spammers will, too, and since OpenAI has access to what they're asking and can easily flag their questions, they can feed misleading guidance to spammers.