From disruption
to ownership.

Every major media disruption followed the same arc. The publishers who survived were not the ones who got the technology right first — they were the ones who stopped pretending the pattern did not apply to them.

Scan your domain free →Read the full picture ↓
THE PROOF

This is not theory.
We scanned 5,125 domains.

The global average AI Act readiness score is 4.4 out of 100. One in four publishers has no AI policy of any kind. Half are allowing AI systems to crawl and train on their content with no strategy, no declaration, and no monetisation. This is the “default gap” — and it is not a technology problem. It is a decision problem.

21%
Block all AI crawlers
47%
Allow all — no strategy
28%
Have set no policy at all
13%
Have published llms.txt
4.4
Average AI Act score
out of 100
99
Countries scanned
Find your score among 5,125 publishers →
THE BRIEFING

Open or Closed: The Decision You Didn’t Know You Were Making

Right now, publishers are making one of the most important decisions about their future — without realising they’re making it.

And in doing so, they’re deciding what the rest of us get to see — and what we don’t.

It’s not happening in strategy meetings. Not in boardrooms. It’s happening in files most people have never heard of: robots.txt, ai.txt, llms.txt, JSON-LD.

I’ve worked with publishers for a long time. And like most people in this industry, I’ve watched the traffic change. It didn’t disappear — people didn’t stop being curious about the world. But it went somewhere else. To AI. People started asking questions and getting answers directly, without ever arriving at a publisher’s website.

I wanted to understand what that means in practice. How do you become the answer? What does it take? And what are publishers around the world actually doing about this right now?

So over the last four months I’ve been doing the research — looking at close to 5,000 publisher domains across 99 countries, analysing whether they have AI policies, how they’re configured, whether their doors are open or closed to AI systems, and whether any of it looks like a deliberate choice or just a default nobody has revisited.

What I found was interesting. And a little uncomfortable.

Open as page →
THE DATA

AI Readiness Global Dashboard

Enter your email to explore the full dataset

Free. No spam.

THE SOLUTION

Adaptation starts
with infrastructure.

Fujifilm survived the digital transition that destroyed Kodak — because they built new infrastructure while the old model still had margin. The window is still open for publishers. Three steps. In this order.

01

Declare your position

robots.txt tells AI crawlers whether they can enter. ai.txt tells them why. llms.txt tells them what matters. Right now, 28% of publishers have said nothing at all — and silence is not neutrality. It is absence. Start with a declaration, even if it is just a block. Undeclared is the worst position.

robots.txt AI policyai.txt declarationllms.txt authority signal
02

Build your signal

An AI system that can reach your content still needs to understand it. JSON-LD structured data, named authorship signals, and editorial policy pages are the infrastructure that turns a crawlable site into a citable source. Without them, you are present but unidentifiable — and AI systems reach for whoever is identifiable.

JSON-LD structured dataNamed editor visibleEditorial policy pageNewsArticle schema
03

Get on the new rails

The first generation of AI monetization infrastructure exists: Tollbit, Cloudflare Pay Per Crawl, the Perplexity Publisher Program. These are not theoretical. They are live, and publishers who qualify are already earning from AI traffic. Qualification requires exactly the infrastructure from steps 01 and 02. The window is open. The pattern says it does not stay open long.

Tollbit integrationPerplexity Publisher ProgramCloudflare Pay Per Crawl

The research series

Three reports. Each one answers a different question publishers are avoiding.

Whitepaper — coming soon

The AI Act Compliance Gap

The global average AI Act readiness score is 4.4 out of 100. What that means for publishers facing a compliance deadline — and what the 4.4% who score above 50 are doing differently.

Notify me when it’s ready →

Whitepaper — coming soon

The Nordic Paradox vs. The World

Norway (67%) and Sweden (66%) are blocking at three times the global average. Denmark leads on formal AI declaration. Five countries, five strategies — and what it means for the future of Nordic journalism in AI systems.

Notify me when it’s ready →

Whitepaper — coming soon

The Licensing & Monetization Framework

Only 13% of publishers have a llms.txt. They are the only ones currently open for business with AI platforms. Tollbit, Cloudflare Pay Per Crawl, Perplexity Publisher Program — what qualifies, what pays, and how to get there.

Notify me when it’s ready →

Where does your domain stand?

Anseri runs the same scan that produced every number in this report. See your score against 5,125 peers — free, in 60 seconds.

Get your free AI readiness scan →Book a strategy call