IANAL, but lately I've had this quixotic daydream of a combination accept-cookies / agree-to-TOS page that comes up, and the Terms of Service says by proceeding they agree to give the site-owner an perpetual, irrevocable, and royalty-free to use and re-license any future content that they create using any generative AI that was trained using the website contents.
Then you carefully log what LLM user-agents/IPs go past that agree, along with some very distinctive secretly crawlable pages which have contents that can be distinctively reproduced back out of the model if needed.
Then whenever SomeShittyLLM posts "articles", everybody with that TOS that was crawled gets to duplicate it without ads for free. :P
Then you carefully log what LLM user-agents/IPs go past that agree, along with some very distinctive secretly crawlable pages which have contents that can be distinctively reproduced back out of the model if needed.
Then whenever SomeShittyLLM posts "articles", everybody with that TOS that was crawled gets to duplicate it without ads for free. :P