# Applebot-Extended: Opt Out of Apple AI Training

> Applebot-Extended is a robots.txt token, not a crawler. Disallow it to keep Apple from training its AI models on your pages. Apple search stays the same.

Applebot-Extended is a robots.txt token from Apple, not a crawler. It tells Apple whether it can train its generative AI models on pages that its crawler already collected. If you do not want your content in that training, disallow Applebot-Extended in robots.txt. Your pages stay in search results.

## What the token controls

The [About Applebot](https://support.apple.com/en-us/119829) help page describes the token. A user agent token is the name that a robots.txt group refers to. The help page calls Applebot-Extended a secondary user agent. With it, you can opt out of the use of your content to train general purpose foundation models. These models power generative AI features, including Apple Intelligence, Services and Developer Tools.

The help page states that Applebot-Extended does not crawl webpages. The token only decides how the company can use data that the main crawler collected. For this reason, you will not see Applebot-Extended in your access logs. Every request comes from [Applebot](/ai-crawler-bot/applebot/), and that guide explains how to verify the requests.

## What the training data includes

A research post from June 2024, [Introducing Apple's On-Device and Server Foundation Models](https://machinelearning.apple.com/research/introducing-apple-foundation-models), describes the training data:

- Licensed data.
- Publicly available data that the web crawler collected. Web publishers have the option to opt out.
- Filters remove personally identifiable information, such as social security and credit card numbers.
- Filters also remove profanity and other low-quality content.

The post also says that Apple never uses the private personal data of its users, or their activity, to train these models.

## Opt out in robots.txt

If you want to opt out for the whole site, add this group to your robots.txt file:

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

If you want to opt out for one folder only, disallow that path. This is the example from the help page:

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

Use this exact name in the group. The help page does not say that rules for other AI tokens, such as [Google-Extended](/ai-crawler-bot/google-extended/), have any effect here.

## Make sure that the rule is correct

A small typing error can change the result:

1. Make sure that the group says `Applebot-Extended`, with the hyphen. A group for the crawler name alone controls crawling and search, not training.
2. Make sure that the file is at the root of each host, for example `https://www.example.com/robots.txt`.
3. Open the file in a browser after you upload it, and make sure that the server returns the new version.

## What the opt-out does not change

A Disallow rule for this token has a narrow effect. The help page documents these limits:

- Pages that disallow the token can still appear in search results in Spotlight, Siri and Safari.
- The rules for this token are not considered in ranking for search.
- The crawler can still fetch the pages. To stop the crawling, write rules for the crawler itself.

The token also does not control the use of your content as context for AI answers in Siri and Search. A robots meta tag, `nosnippet`, controls that use. The [Applebot guide](/ai-crawler-bot/applebot/) explains it.

## Why sites allow or block it

The help page gives a reason to allow the token: "Allowing Applebot-Extended will help improve the capabilities and quality of Apple's generative AI models over time."

Many large sites chose to block it. In 2024, [Wired reported](https://www.wired.com/story/applebot-extended-apple-ai-scraping/) that major publishers and platforms, such as The New York Times and Facebook, had opted out. Wired found this in the public robots.txt files of these sites.

## Open questions

The help page does not say when an opt-out takes effect. It also does not say whether content that was collected before you added the rule stays in training data. If these points matter to you, write to applebot@apple.com, the contact address on the help page.

---

[View the canonical page](https://aichatwatch.com/ai-crawler-bot/applebot-extended/) · [Browse llms.txt](https://aichatwatch.com/llms.txt)