> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.ada.cx/docs/setup/redactions/redaction-settings/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.ada.cx/_mcp/server.

# Redaction settings

## Overview

You can ensure that your end users' information stays safe and secure by telling your AI Agent what types of information it should keep private. That way, if end users provide high-risk information that you do not want your AI Agent to use, you can ensure it does not leave the conversation.

## Use cases

Redactions enable the following scenarios:

* **Protecting sensitive identifiers**: Prevent IDs, account numbers, or other high-risk identifiers from being stored in conversation transcripts or used by your AI Agent.
* **Securing financial information**: Redact bank account details or other payment information that end users may inadvertently share.
* **Compliance with data regulations**: Support privacy requirements by ensuring specific data types are not retained in conversation history or voice recordings.

## Capabilities & configuration\[#how-redactions-work]

Redactions prevent sensitive information from being used or stored by your AI Agent.

When Ada redacts information:

* Your AI Agent does not use redacted information in its reasoning or when generating responses for your end users. It acts as if the end user never provided that information in the first place.
* If your AI Agent hands off the conversation to a human agent, the human agent is unable to see the redacted information.
* Redacted information does not appear in conversation transcripts, and is covered with beeps in voice recordings.

> **Note**
>
> Be careful not to redact information that your AI Agent needs to resolve inquiries. Over-redacting can impact its ability to help end users.

You can control redactions by using regular expressions (regex) to define exact matching patterns your AI Agent will follow.

* Because regex is very literal, if values do not follow a consistent pattern (e.g., names or addresses), your AI Agent cannot reliably redact them.
* If you need help writing your regex patterns, try a site like [RegExr](https://regexr.com/) to help you put it together and test that it works.

> **Tip**
>
> It can be tricky getting your regex patterns to work as intended. They can often be too exact, excluding some use cases; or too broad, including more than you wanted them to. Make sure you test a variety of use cases to ensure you are capturing the right amount of information.

### Redactions and images

Redactions also protect the text your AI Agent receives when an end user sends an [image](/docs/other/images).

* The Agent receives a text description of each image, not the image itself.
* Redactions run on that description before the Agent reasons over it.
* A matched pattern disappears from the Agent's reasoning and from the transcript, the same as for typed text.
* Redactions do not alter the image file.
* The Agent detects images that show a payment card, an identity document, or another credential, and withholds them independent of your redaction rules.
* Detection is best-effort. Keep redaction rules for identifiers that must not appear in transcripts.

## Quick start

Create a custom Redaction in a few steps.

On the Ada dashboard, go to **Config > AI AGENT > Settings**, then open the **Preferences** tab.

Under **Redactions**, click **Create redaction**.

Enter a **Redaction name** and **Redaction regex** pattern.

Test your regex with example values, then click **Save**.

For detailed instructions, see [Implementation & usage](#create).

## Implementation & usage

Create and manage custom redactions for your AI Agent.

### Create a custom Redaction\[#create]

Define a regex pattern to redact specific types of sensitive information.

**To create a custom redaction:**

1. On the Ada dashboard, go to **Config > AI AGENT > Settings**.

2. On the **AI Agent Setup** page, on the **Preferences** tab, under **Redactions**, either enable an existing custom redaction, or click **Create redaction** to create a new one.

   The **Create redaction** window opens.

3. Under **Redaction name**, enter a name for your redaction. Your AI Agent does not send the redaction name to the LLM; this is for your use only.

4. Under **Redaction regex**, enter your regex pattern.

5. Under **Test your regex**, enter an example value to ensure that your regex pattern is working properly. Make sure you test values you want your AI Agent to redact, as well as values you want it to leave alone.

6. Click **Save**.

## Related features

* [Personality setup](/docs/setup/persona/personality-setup): Configure your AI Agent's identity and communication style.
* [Data retention](/docs/other/data-retention): Manage how long conversation data is stored.
* [Images](/docs/other/images): Learn how your AI Agent reads images and which images it withholds.

---

Have any questions? Contact your Ada team, or email us at [](mailto:help@ada.cx?subject=Help%20Docs%20inquiry).