Research.
We study how AI reads the web and what it means for the people who make content. Our reports use public data you can check yourself.
Websites are locking the wrong door.
We read the robots.txtrobots.txt: The file on a website that tells crawlers which pages they may fetch. file of 457 news websites. Most block the robots that learn from old stories. Few block the robot that writes quick answers and keeps the reader.
- 457
- News websites checked, in 12 groups, on June 26, 2026.
- 4 in 5
- Let the answer robot in.
- 50.8%
- Of the sites we could grade got an F.
Notes from the standards work.
From our seat on the C2PA Text ProvenanceText provenance: A signed label for text: who signed it, when, and whether the words changed. Task Force.

- ISSMAD 2026
- Beyond the Visual Bias. Text has no pixels. WatermarksWatermark: A mark placed in the content itself, visible or hidden, so it can be found again after the file is copied., signed C2PAC2PA: The Coalition for Content Provenance and Authenticity: the group that publishes the open standard for content labels. labels and outside checks each answer a different question.
- Research note
- The Short-Text Provenance Gap. New EU rules admit that marking short text is hard. Here is why that gap in provenanceContent provenance: The record of where a piece of content came from and how it changed, carried with the content itself. matters, and how we close it.
Coming soon: The Future of IP and Media.
How rights, licensing and labels reshape the media business as AI reads the web first. Join the list and we send it the day it is out.
Create. Mark. Endure.
See which AI robots your own site lets in. It takes one minute and needs no signup.