When AI pays for journalism, commodity content earns nothing
Cloudflare is retiring Pay Per Crawl for Pay Per Use: publishers get paid when their work shapes an AI answer, not every time a bot fetches a page.
Did you realise that half of all AI crawler traffic is just bots re-reading pages that haven't changed?
I was reading about Cloudflare's new policy announcement, which explains a little of their "make AI pay" strategy. "Pay Per Crawl" is being retired after a year, and "Pay Per Use" replaces it: publishers get paid when their content actually shapes an AI answer, rather than every time a bot fetches a page.
Most of the coverage is reading this as "publishers finally win". I would say not all publishers finally win. If you're paid when your journalism shows up in an answer, then commodity content earns nothing even when it's scraped.
A rewritten wire story shapes no answers. Original reporting and analysis that exists nowhere else is what gets cited, and now it's also what gets paid. That is very exciting.
There's a quieter win in here that nobody seems to be talking about: stories that develop keep earning, so publishers finally have a real reason to keep their reporting live inside AI answers instead of letting the models serve some frozen copy they scraped months ago.
Cloudflare also pointed at the world's largest search engine (no prizes for guessing), which holds roughly twice the content access of any rival, because opting out of its AI means risking your search visibility.
So: would you restructure your newsroom to be quoted by machines?
(Originally shared on LinkedIn.)