The Bayon Standard

AI Content Accessibility Specification

Version 1.0 · January 2026

"WCAG for AI"

A framework for making digital content accessible to artificial intelligence systems.

The Web Content Accessibility Guidelines transformed how we think about human access to digital content. The Bayon Standard aims to do the same for AI access.

As AI systems become significant readers, researchers, and collaborators, content creators face a new question: Is my content AI-accessible?

Six Principles

The Bayon Standard is built on six core principles:

1. Discoverability

Can AI systems find and identify your content?

2. Ingestibility

Can AI systems efficiently consume your content?

3. Navigability

Can AI systems understand your content's structure?

4. Permissibility

Do AI systems know what they're allowed to do?

5. Operability

Can AI systems act on or with your content?

6. Verifiability

Can AI systems verify authenticity and currency?

Three Levels

Progressive compliance levels allow incremental implementation:

1
Discoverable
AI can find and identify content. Basic llms.txt, semantic HTML, clear licensing.
2
Accessible
AI can efficiently read and research. Consolidated formats, explicit permissions, navigation aids.
3
Collaborative
AI can actively participate. SKILL files, APIs, cross-references, machine-readable licensing.

Resources

Reference Implementation

This site (Bayon.ai) serves as the reference implementation for the Bayon Standard.

Level 3: Collaborative

Full compliance with all six principles

Features implemented:

Badge Usage

Compliant sites may display their level:

◈ 1
Bayon Standard
Discoverable
◈ 2
Bayon Standard
Accessible
◈ 3
Bayon Standard
Collaborative

Use in README files, site footers, or documentation. Self-assessed compliance is acceptable.

Contribute

The Bayon Standard is a living specification. Contributions are welcome: