Is ai.txt an established web standard?

Definition: ai.txt is a filename used by separately authored proposals for machine-readable policies about AI use of website content. As of 17 October 2026, the AI.TXT Internet-Draft is an individual proposal without formal standing in the IETF standards process; Spawning publishes its own generator and training-permission convention.

Key facts

• The IETF Datatracker record identifies Kayla Cardillo's AI.TXT document as an active individual Internet-Draft.
• Cardillo's draft text states an intended Informational status and proposes /.well-known/ai.txt with an ai.json companion.
• Spawning's generator documentation describes permissions for commercial AI training and defaults to opting all content out.
• The Robots Exclusion Protocol, which defines robots.txt, is published as the Internet Standards Track document RFC 9309.

What is the AI.TXT draft's publication status?

An Internet-Draft is a working document that can change or be replaced. Cardillo's revision draft-car-ai-txt-wellknown-00, dated 12 June 2026, carries an expiry date of 14 December 2026. Its status record explicitly states that the IETF has not endorsed it and that it has no formal standing in the standards process.

The draft's intended Informational status describes its publication aim. The document remains a draft. Its request to register the ai.txt and ai.json well-known URI suffixes is part of that proposal. A well-known URI is a resource address under the /.well-known/ path.

What does Spawning publish?

Spawning's announcement of 30 May 2023 describes a file at the website root for communicating content and media permissions. The company presents it as a practical declaration mechanism and says it intends to integrate other emerging conventions into its tooling.

The generator's published page data documents toggles for allowing or blocking training use by content type. All content starts opted out. Selecting allow communicates permission for the selected media type.

These documents have different authors and scopes. Cardillo's draft acknowledges Spawning's earlier use of the filename and proposes broader fields covering indexing, caching and licensing. Identifying the author and format is therefore necessary when discussing an ai.txt implementation.

What can a website operator configure?

Cardillo's format proposal specifies a text file at https://example.com/.well-known/ai.txt. Its minimal example reads:

# ai.txt
Spec-Version: 1.0
Site-Name: My Blog
Site-URL: https://myblog.com
Training: deny

This is the draft's illustrative configuration. Training: deny declares a training policy. The draft treats the file as advisory and requires servers to enforce access controls and rate limits independently. Its suggested agent behaviour belongs to the proposal, rather than a documented commitment by an answer engine.

For crawling permissions, the published Robots Exclusion Protocol defines the separate /robots.txt resource. The AI.TXT draft explicitly allows both files to coexist.

Related terms

• ai.txt proposals
• robots.txt
• TDM Reservation Protocol

FAQ

Does an IETF-hosted draft establish a standard?

No. The Datatracker says anyone may submit an Internet-Draft and explicitly identifies this submission's lack of IETF endorsement.

How can I check a Spawning installation?

The generator's installation answer tells you to open yourdomain.com/ai.txt, substituting your domain, and check that the file appears as plain text.

Does publishing the draft's file enforce access restrictions?

The draft's compliance section describes advisory declarations. You need server-side controls to enforce access restrictions independently of that file.

Sources

1. IETF Datatracker, "AI.TXT: A Declaration File for AI Usage Preferences, Licensing, and Policy", updated 12 June 2026
2. Kayla Cardillo, "AI.TXT: A Declaration File for AI Usage Preferences, Licensing, and Policy", 12 June 2026
3. Spawning, "Spawning's ai.txt Generator", undated, documentation page data
4. Spawning, "ai.txt: A new way for websites to set permissions for AI", 30 May 2023
5. IETF, "Robots Exclusion Protocol", September 2022

---

Check how a page reads to answer engines. Open the GEO score analyzer