The open standards ContentSignals works with

robots.txt

The Robots Exclusion Protocol, first defined in 1994, lets a site tell crawlers which parts of it they may visit. The IETF published it as RFC 9309, a Proposed Standard, in September 2022. The standard is explicit that its rules "are not a form of access authorization".^⁠[1]

In the model A site's robots.txt is the first thing its owner says. Where a crawler may go sets the outer edge for every other signal.

Cloudflare's Content Signals Policy

Cloudflare published the Content Signals Policy on 24 September 2025 under CC0, which allows anyone to implement and use it freely.^⁠[2] It adds a Content-Signal line to robots.txt that says how content may be used after it has been accessed, for three uses:^⁠[2]

search

"building a search index and providing search results", which the policy says does not include AI-generated search summaries

ai-input

"inputting content into one or more AI models", for example to ground an answer

ai-train

"training or fine-tuning AI models"

A yes grants a use, a no withholds it, and a signal left out neither grants nor restricts it. The values express preferences and are not technical barriers against scraping.^⁠[2]

In the model Each value is a signal from the owner, kept with the content in the graph. A no for AI training is part of what a piece is, just as its citations are.

This site's own robots.txt carries the line Content-Signal: search=yes, ai-input=yes, ai-train=yes, because its pages are written to be found and read by search and AI engines alike.

AI Preferences at the IETF

The IETF's AI Preferences working group will standardise "building blocks that allow for the expression of preferences about how content is collected and processed": a common vocabulary, and ways to attach it to content through robots.txt and HTTP response header fields.^⁠[3]

In the model When that vocabulary is published, ContentSignals reads it the same way: as the owner's own signal, carried with the content.

IndexNow

IndexNow is "an easy way for websites owners to instantly inform search engines" that a page has changed,^⁠[4] whether it was added, updated or deleted. Search engines that adopt the protocol share the URLs they are sent with all the other participating engines.^⁠[5] The site names Microsoft Bing, Naver, Seznam.cz, Yandex and Yep as supporting it.^⁠[4]

In the model A new or changed page is a signal too. IndexNow is how a site gives that signal straight away, instead of waiting for a crawler to notice.

schema.org

Schema.org was founded by Google, Microsoft, Yahoo and Yandex, with the mission to create, maintain and promote schemas for structured data on the internet. Since April 2015 its main forum has been a W3C community group.^⁠[6]

Its vocabulary already names many signals. An InteractionCounter records a type of action, such as a like, a share or a comment, and how many times it happened;^⁠[7] AggregateRating gives "the average rating based on multiple ratings or reviews";^⁠[8] and the citation property records a reference to another work.^⁠[9]

In the model Structured data lets a page state its own signals in a form any engine can read. Every page of this site carries schema.org data, including the list of sources it cites.

What building on them means

ContentSignals adds no new tag that a site has to adopt. It reads what open standards already let a site say, and follows it through the graph, valuation and points. Each standard keeps its own publisher, licence and rules, and naming it here claims no relationship with the people who publish it.