Skip to content
← All articles

On This Day in AI · 1 July 2026

The Day the Web Started Charging AI to Read It

On this day, 2025, Cloudflare flipped one default and quietly changed who controls the open web: AI bots could no longer help themselves to your website for free.

July 1, 2025: Cloudflare starts blocking AI crawlers by default and lets websites charge them for access


On this day, 2025, Cloudflare flipped one default and quietly changed who controls the open web: AI bots could no longer help themselves to your website for free.

Back then, the deal had been silent and one-sided for years. AI companies sent out bots that copied huge swathes of the internet, your site included, to train their models, and nobody asked you first. If you ran a business with a website, a blog, or a shop, your words and your photos were feeding tools you never agreed to. It felt like there was nothing you could do, and most people assumed that was just the price of being online now.

Here is what actually changed. Cloudflare sits in front of about a fifth of all websites, like a doorman for a huge stretch of the internet. On this day it changed the default from “let every AI bot walk in” to “stop and ask first.” Then it added a third choice beyond yes and no: charge. A site could now hand an AI bot a bill, an actual “payment required” notice, before letting it read a single line. Allow, block, or charge. Your call.

So what has happened in the year since? This stopped being a stunt and turned into plumbing. Stack Overflow wired up the same charge to crawl model for its answers. Cloudflare folded the whole thing into one simple console available on every plan. And the numbers tell the story on their own: websites now send more than a billion “payment required” replies to AI bots every single day. The era of scrape everything and ask nobody is closing. Without much noise, the question flipped from “can they take it” to “what will they pay for it,” and the people who own the content ended up with leverage almost nobody expected them to have.

That is the pattern worth seeing. Last summer’s frightening headline, that AI was going to eat the open web, turned out to be the start of the web learning to charge rent. The takeaway is not which setting to switch on today. It is that the things which feel powerless and permanent in AI rarely stay that way for long. Give it a year, and the giant can be the one being asked to pay at the door.

If you want to read more about the news on this day: The announcement: Content Independence Day: no AI crawl without compensation How the charging works: Introducing pay per crawl

Yesterday’s breakthrough is today’s normal.

The newsletter

Get On This Day in AI every week.

One story from AI's recent past, what happened next, and what it means for running a small business. Five minutes, no hype.

Free 30 minute call

Tell us what's eating your week.

You leave with two or three ways to save time, whether we work together or not.