# Opt-Out Tokens: Decline AI Training, Keep Search

> Opt-out tokens decline AI training per operator without leaving search: Google-Extended, Applebot-Extended and noai opt out of training, not search.

## Opt-out tokens decline AI training per operator

Opt-out tokens let you decline AI *training* on a per-operator basis without leaving search. `Google-Extended` is a robots.txt user-agent token that, as reported by Google, opts a site out of training Gemini and Vertex AI generative models while keeping normal Google Search indexing. `Applebot-Extended` does the equivalent for Apple Intelligence. Both are operator-controlled and honored at the operator's discretion, so verify exact behavior against each operator's primary docs at build (Google Search Central, Apple's Applebot docs).

```
User-agent: Google-Extended
Disallow: /

User-agent: Applebot-Extended
Disallow: /
```

## The training-versus-search distinction is load-bearing

A **training opt-out is not a search block**. Disallowing `Google-Extended` or `Applebot-Extended` stops generative-model training but leaves the retrieval crawler free to index and cite you. Blocking the *search* crawler instead removes you from results — and from the AI answers built on them — forfeiting citations and referrals. Most publishers should refuse training while keeping the retrieval crawler welcome.

## noai and noimageai signal content-level refusal

The `noai` and `noimageai` meta directives signal at the page level that text and images should not be used for AI. They are adoption-dependent — honored at operator discretion, with no edge enforcement. A statement of intent, not a hard control.

```
<meta name="robots" content="noai, noimageai">
```

## robots.txt remains the per-crawler opt-out of record

For a hard, per-crawler refusal, a `User-agent` + `Disallow` rule (e.g. naming `GPTBot` or `ClaudeBot`) remains the opt-out of record — still compliance-based, but the long-standing convention. Per-bot records live in the registry.

Related: [Google-Extended defined](/glossary/google-extended) · [robots.txt-AI spec](/protocols/licensing/content-signal) · [crawler registry](/crawlers) · [should you block AI?](/access-economics/should-you-block-ai) · [AI access economics](/access-economics)