> ## Documentation Index
> Fetch the complete documentation index at: https://www.plain.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# AI and your data

> Which AI providers Plain sends data to, what they can do with it, and how to turn AI features off.

When Plain's AI features are on, thread content is sent to OpenAI and Anthropic to generate a response. Neither provider uses your data to train its models, and you can turn every AI feature off from your workspace settings.

## Which providers Plain uses

Plain uses OpenAI and Anthropic to power its AI features, which include [Ari](/docs/product/agents/ari), [Sidekick](/docs/product/agents/sidekick), thread summaries, [suggested responses](/docs/product/agents/ari/suggested-responses), [AI triage](/docs/product/platform/ai-triage), and [AI slash commands](/docs/product/platform/ai-slash-commands).

Both providers are listed as sub-processors in Plain's [Data Processing Agreement (DPA)](https://www.plain.com/legal/dpa), which sets out the full terms Plain processes data under.

## Zero data retention

Plain has opted into [OpenAI's zero data retention program](https://developers.openai.com/api/docs/guides/your-data#zero_data_retention). Data sent to OpenAI is not stored once the API call completes, and OpenAI does not use it to train its models.

Anthropic does not use data sent through its API to train its models either.

## Turning AI features off

AI features are on by default. A workspace admin can turn them off at any time:

<Steps>
  <Step title="Open your AI settings">
    Go to **Settings → Plain AI → Configuration**.
  </Step>

  <Step title="Turn off the features you don't want">
    Each feature can be disabled on its own, or you can disable AI entirely.
  </Step>
</Steps>

Changes take effect immediately. Once a feature is disabled, no thread content from it reaches any AI provider.