We just got back from a week in Cannes (flex, we know), and we want to tell you about the single most interesting thing we heard the entire time we were there.
It wasn't about AI, exactly. And it wasn't about creators, exactly. But it was kinda about both, and when we were thinking about our time there and “takeaways” it definitely stood out as the most compelling thing to share.
We were sitting in on a panel with Matthew Prince and Tyler Denk, the CEOs of Cloudflare and beehiiv, and off the back of an announcement the two of them just made together, Matthew laid out a theory about how the economics of the entire internet might be about to change.
A quick detour: what even is Cloudflare?
If you've never had to think about Cloudflare, congratulations, that means it's doing its job!
Picture the internet as a city. Cloudflare is a big chunk of the plumbing, the roads, and the security guards you never notice until something breaks. When you visit a website, there's a decent chance your request passes through Cloudflare's pipes first, which speed the site up, block the bad actors trying to knock it over, and generally keep the lights on. A meaningful slice of all web traffic runs through them (which is why their market cap is $87 billion as of writing this)
The part that matters for today is that because they sit in that position, Cloudflare can see and control what reaches a website, including the AI bots that quietly crawl the web to feed the models. They're the bouncer at the door, and they can see exactly who's trying to walk in - and check IDs!
What they announced
Cloudflare and beehiiv (the platform a lot of creators publish on, 135,000-odd publishers and counting) announced they're building Cloudflare's "AI Crawl Control" directly into beehiiv.
In plain terms, every publisher gets a simple choice, presented as two doors.
🚪 Door one: throw it wide open. Let the AI crawlers take your work freely, on the bet that showing up inside AI search and AI answers sends you discovery and distribution you couldn't buy otherwise.
🚪🚪 Door two: lock it down. Block the scrapers, protect your archive, and hold your content back for future monetization and licensing deals.
You also get a dashboard showing which AI bots are knocking, which ones you're turning away, and what traffic they send back to you. While this sounds like a small feature, it’s pointing at the future of a very big idea.
What Cloudflare was actually saying
For about twenty years, the internet has run on one basic bargain. Content gets made, content gets viewed, and that attention gets sold to advertisers. Traffic is the currency. That single loop has funded Google search / ads, blogs, journalism, YouTube, and a huge portion of the creator economy, including the part where a brand pays to sit next to your audience.
Widespread LLM adoption breaks that loop. When someone asks ChatGPT a question instead of clicking through ten blue links, the model becomes the interface. It reads your article, digests it, and hands the answer straight to the person. They got what they came for, and they never landed on your page. Which means the ad you sold against that page shows to nobody, and the pageview that used to pay for the work simply never happens.
So the engine that funded the internet starts to sputter because the traffic got rerouted through a machine before it ever reached your door.
Cloudflare's bet, is that if the LLM is the one consuming the work, the LLM should be the one paying for it. The people actually making the stuff (publishers, creators, journalists) should be compensated by the models feeding on it, rather than by an ad impression that increasingly reaches no one. And Cloudflare, sitting on all that plumbing, wants to build the toll booth that makes it possible.
That's a vert different shape for the web: a world where giving content away stops being the only move.
Discovery… or monetization?
Whether that future helps you or hurts you depends entirely on how your business makes money, and it splits creators into two camps (or maybe forces people to create two separate streams of content?)
For some, content is a billboard. The revenue generated lives one step past it, in a book, a membership, a service, a course. If that's you, an LLM chewing through your free stuff and sending people back toward you is distribution you'd otherwise have to pay for. You want the doors flung wide open, because discovery feeds the thing that actually makes you money, a book, a membership, a course.
For others, the content is the product. The archive itself is the asset, the exact thing people came to you (and paid) to get. If a model swallows it and serves your ideas back up with no name attached and no check in the mail…you don’t have a business anymore. If that's you, blocking the scrapers and making the models pay is critical to the longevity of the biz.
So, where does that leave you
You don't need to do anything about it today. There's no framework to download, no five steps to implement before Friday. We'd just suggest noticing which of those two doors your business would want to walk through, and why.
.webp)
