---
title: "AI Terms"
description: "Publish an explicit AI Terms policy for search, answers, training, and agents."
canonical_url: "https://nuxtseo.com/docs/ai-ready/guides/ai-terms"
---

> Index of every page on this site: [llms.txt](https://nuxtseo.com/llms.txt)

Set `aiReady.aiTerms`{lang="ts"} to publish your policy at `/.well-known/ai-terms`{lang="http"}.
The feature follows the [AI Terms 0.1 draft](https://aipaygen.com/ai-terms).
It stays disabled until you configure a policy.

Set `site.url`{lang="ts"} and give all four uses an access rule.
Set `default`{lang="ts"} for clients that omit the `AI-Purpose`{lang="http"} header.
Set `updated`{lang="ts"} to the date you last changed the policy, in `YYYY-MM-DD` format.

```ts [nuxt.config.ts]
export default defineNuxtConfig({
  modules: ['nuxt-ai-ready'],
  site: { url: 'https://example.com' },
  aiReady: {
    aiTerms: {
      updated: '2026-11-18',
      default: 'deny',
      uses: {
        search: { access: 'allow', attribution: 'required' },
        answer: { access: 'allow', attribution: 'required' },
        training: { access: 'deny' },
        agent: { access: 'allow' },
      },
    },
  },
})
```

After you start the server, substitute your site URL and fetch the policy.
Check that the response contains your rules and that `site` matches your site's origin.

```sh
curl -i https://example.com/.well-known/ai-terms
```

The endpoint supports `GET`{lang="http"}, `HEAD`{lang="http"}, and cross-origin discovery.
Check a page's response for a `Link`{lang="http"} header with `rel="ai-terms"`.
Check `robots.txt`{lang="txt"} for an `AI-Terms`{lang="txt"} directive.
Prerendering writes a static policy file.
AI Terms also works with `database: false`{lang="ts"}.

With `app.baseURL: '/docs/'`{lang="ts"}, the discovery URL becomes `https://example.com/docs/.well-known/ai-terms`.
Nuxt redirects origin-root discovery requests to that route.
The document's `site`{lang="ts"} remains `https://example.com`, because the draft describes an origin policy.
On static hosting, configure that origin-root redirect if your host does not provide it.
Nuxt Robots disables `robots.txt` generation for apps with a base URL.
Add the `AI-Terms` directive to the origin's own `robots.txt` file in that case.

AI Terms publishes a statement of your policy.
Your server or gateway must enforce access rules, verify client identities, collect payments, and issue receipts.
Existing `robots.txt`{lang="http"} rules and [content signals](https://nuxtseo.com/docs/ai-ready/guides/content-signals) still apply.
Publishing an `allow`{lang="ts"} rule does not override those restrictions.

## Uses and path overrides

| Use        | Purpose                                            |
| ---------- | -------------------------------------------------- |
| `search`   | Index pages for AI search links and snippets       |
| `answer`   | Fetch a page to answer a person's current question |
| `training` | Collect content for model training or fine-tuning  |
| `agent`    | Read pages or call endpoints to complete a task    |

Each use requires `access`{lang="ts"}: `allow`{lang="ts"}, `deny`{lang="ts"}, or `paid`{lang="ts"}.
Optional `attribution`{lang="ts"} accepts `required`{lang="ts"}, `requested`{lang="ts"}, or `none`{lang="ts"}.
Use `paths`{lang="ts"} to override selected uses:

```ts
paths: [
  {
    match: '/members/*',
    uses: {
      search: { access: 'deny' },
      answer: { access: 'deny' },
    },
  },
]
```

In the draft, `*` matches within one path segment.
The most specific matching entry wins and overrides only the uses it lists.
Configure Nuxt route rules separately if your site needs them.

## Describe an existing payment setup

If your gateway already enforces x402 payments, describe its price and endpoint:

```ts
export default defineNuxtConfig({
  modules: ['nuxt-ai-ready'],
  site: { url: 'https://example.com' },
  aiReady: {
    aiTerms: {
      updated: '2026-11-18',
      default: 'deny',
      uses: {
        search: { access: 'allow' },
        answer: { access: 'allow' },
        training: { access: 'deny' },
        agent: {
          access: 'paid',
          price: { amount: '0.002', currency: 'USD', per: 'page' },
        },
      },
      payment: {
        protocol: 'x402',
        version: 2,
        endpoint: 'https://pay.example.com/x402',
        network: 'eip155:8453',
        asset: 'USDC',
      },
    },
  },
})
```

A paid use needs a decimal string price and an absolute HTTPS payment endpoint.
Use an endpoint URL without credentials, query parameters, or a fragment.
Path overrides can provide their own endpoint or inherit the top-level `payment`{lang="ts"}.
A paid `default`{lang="ts"} requires a top-level endpoint.
The module rejects invalid configuration during Nuxt setup.

The configuration also accepts the draft's `identity`{lang="ts"}, `receipts`{lang="ts"}, `contact`{lang="ts"}, `report_only`{lang="ts"}, and `preview`{lang="ts"} fields.
Paid and denied uses can declare a verified client's `free`{lang="ts"} allowance.
Publish these fields only when they describe your server's behavior.
Set `aiTerms: false`{lang="ts"} to disable an inherited policy.

## Sitemap

See the full [sitemap](https://nuxtseo.com/sitemap.md) for all pages.
