Objective
Detection needs tuning as the traffic you see changes. This article shows you how to open one of your own rules and change its conditions, confidence score, or outcome.
Product
Fraud Risk Engine
Environment
Twilio Console
User Account Permission/Role(s) Required
Your user needs permission to update Fraud Risk Engine rules. Without it, Twilio reports an authorization error instead of saving.
Procedure
- In the Twilio Console, select Identity & security in the left sidebar.
- Select Fraud Risk Engine, then Rule Management.
- Find the rule you want to change. Its Rule origin must show that you created it, Twilio's own rules cannot be edited in place.
- Open the rule's actions menu and select View & Manage. The rule opens in an editable panel headed Edit Rule.
- Change what you need:
- Under Rule details, update the Rule name, Event Type, Custom Type ID, or Description. Rule origin stays Custom and is read-only.
- Under Rule configuration, adjust the Conditions, the Confidence Score, and the decision under Actions / Outcome.
- Use the Active switch in the panel header to control whether the rule evaluates traffic.
- Select Save.
- If the Active switch is off, Twilio asks Are you sure you want to save this rule without activating it? Select Save without activating to confirm, or Cancel to go back and activate it first.
Additional Information
- Only rules you created can be edited. Selecting View & Manage on a Twilio-managed rule opens it read-only, clone it first if you need to change it.
- The same validation rules apply as when creating a rule: Rule name must start with a letter and use only letters, numbers, or underscores; Description is capped at 500 characters; and Confidence Score must fall inside the allowed range.
- A saved-but-inactive rule stays in your list with a Deactivated status and does not evaluate any traffic until you turn it on.
- Use the configuration toggle to read the conditions as code instead of building them visually, with a copy control for the code view.
This content was generated by AI and reviewed, edited, and verified by a human.