There already is an OpenAI API to detect texts made by bots. I see a browser extension which tells if selected text is GPT bullshit and also marks the whole page as infested and shares that data across all userbase.