This is a copy-ready ai.txt starter template you can adapt and deploy at the root of your site to declare AI access intent, attribution preferences, and content license. It is a single, lightweight signal that complements robots.txt and llms.txt rather than replacing either.

TL;DR

Deploy ai.txt at https://yoursite.com/ai.txt alongside robots.txt and llms.txt. Use the Balanced variant below if unsure (block training, allow retrieval, require citation, require attribution). As of April 2026 there is no single universally-adopted ai.txt directive set, so treat field names like AI-Training, AI-Retrieval, and AI-Citation as a community convention rather than a ratified spec.

For the broader picture, see the Technical hub.

A note on the ai.txt landscape

Important: Spawning's original ai.txt spec is focused on opt-out signals for media training. The expanded field set used in this template (AI-Training, AI-Retrieval, AI-Citation, Attribution-Required) is a community convention popular among publishers and SEO practitioners; it is not yet ratified by a standards body. Treat it as a clear, machine-readable declaration of intent that AI vendors can choose to honor. Pair it with robots.txt (search-crawler control) and llms.txt (curated content map for LLMs) for the broadest coverage today.

Basic Template

txt

ai.txt for [Your Site]

https://yoursite.com/ai.txt

Last updated: 2026-05-03

Organization

Organization: Your Organization Name

Contact: ai@yoursite.com

Website: https://yoursite.com

AI Access Policy (community convention)

AI-Training: Disallow

AI-Retrieval: Allow

AI-Citation: Required

Attribution

Preferred-Attribution: "Your Site Name (yoursite.com)"

Attribution-Required: Yes

Content License

License: CC BY 4.0

License-URL: https://creativecommons.org/licenses/by/4.0/

Contact for AI-related queries

AI-Contact: ai@yoursite.com

Policy Options Explained

These fields are non-binding as of April 2026: enforcement depends on the AI vendor honoring them. Treat the file as a clear public declaration of intent and a basis for legal/community recourse if violated.

Template Variations

Open Access (Encourage AI Use)

txt

AI-Training: Allow

AI-Retrieval: Allow

AI-Citation: Preferred

Attribution-Required: No

Use when broad reach matters more than control — e.g., open-source documentation, public-interest publications.

Restrictive (Protect Content)

txt

AI-Training: Disallow

AI-Retrieval: Disallow

AI-Citation: Required

Attribution-Required: Yes

Use when content is paid, proprietary, or core to the business model. Pair with robots.txt blocks for known AI training crawlers.

Balanced (Most Common)

txt

AI-Training: Disallow

AI-Retrieval: Allow

AI-Citation: Required

Attribution-Required: Yes

Use for marketing sites, blogs, and B2B content where AI Overviews citation is desirable but training opt-out is preferred.

Deployment

1. Save as ai.txt (plain text, UTF-8, no BOM).
2. Upload to your site root: https://yoursite.com/ai.txt.
3. Verify by visiting the URL in a browser — it must serve as text/plain with HTTP 200.
4. Reference from robots.txt if desired, e.g., # AI policy: https://yoursite.com/ai.txt.
5. Pair with llms.txt for an LLM-friendly content map.
6. Re-publish whenever policy or contact info changes; update the Last updated comment line.

Common Mistakes

• Treating ai.txt as enforceable. It is a declaration of intent. Expect partial honoring, not guaranteed compliance.
• Skipping robots.txt. Crawlers that ignore ai.txt may still respect robots.txt; block AI training UAs there too if you want defense in depth.
• Conflicting llms.txt. If you publish llms.txt with a curated allow-list, make sure ai.txt directives do not contradict it.
• No contact line. AI vendors with policy questions need somewhere to ask; an unmonitored alias guarantees friction.

FAQ

Q: Is ai.txt an official standard?

A: Not as of April 2026. Spawning publishes a media-focused ai.txt format, and the broader field set used in templates like this one is a community convention adopted by publishers. Several efforts are underway to standardize an AI Crawling Protocol; until then, ai.txt is a clear public declaration rather than an enforceable spec.

Q: Do I still need robots.txt?

A: Yes. Robots.txt remains the most-honored crawler-control file. Use ai.txt to declare AI-specific intent and robots.txt to block specific AI crawler user agents you want to deny.

Q: Should I block AI training crawlers entirely?

A: Depends on your business. Publishers monetizing content typically block training while allowing retrieval. Documentation sites usually allow both because citation drives qualified traffic.

Q: Where does llms.txt fit in?

A: llms.txt is a curated content map that helps LLMs find your most authoritative pages. ai.txt declares whether AI can use your content; llms.txt declares what to look at. Most sites benefit from publishing both.

Q: What happens if an AI vendor ignores my ai.txt?

A: Document the violation (timestamps, source URLs, AI output), contact the vendor's policy team, and reference your published ai.txt as evidence of declared intent. Some jurisdictions are starting to treat clearly-published opt-outs as legally meaningful.