Overview

DataDome assigns a Trust Score to every AI agent observed sending traffic to your platform. The Trust Score is a dynamic, continuously updated value between 0 and 100 that reflects how likely an agent is to generate legitimate traffic. It is computed specifically per customer, ensuring that the score reflects the actual behavior observed on your online resources.

How the Trust Score Is Calculated

The Trust Score combines three dimensions:

Dimension	Description
Identification Strength	How securely the agent's identity can be authenticated
Behavioral Intent	The proportion of legitimate vs. fraudulent activity observed, both on your platform and across the DataDome network
Reputation	How well-known the agent is globally, which drives confidence in the score

How the Trust Score Is Used

The Trust Score gives customers an immediate, reliable signal of how likely an agent is to send legitimate traffic — or conversely, how likely it is to be a vehicle for fraudulent activity.

Customers can use it in two ways:

As a decision-support tool: when reviewing and potentially overriding the DataDome-recommended policy for an agent, the Trust Score provides the context needed to make an informed choice. A high score signals that the agent can confidently be granted broad access; a low score suggests applying more restrictive controls, particularly on sensitive endpoints.
As a risk exposure indicator: the Trust Score lets customers assess at a glance the impact of a given agent on their overall exposure to online threats, without needing to dive into raw traffic data.

The Trust Score is visible in the Access Control dashboard alongside the Identification Strength, and is also available as a filtering and grouping dimension in the Explore dashboard.

Note: regardless of the Trust Score or the policy applied to an agent, DataDome continues to analyze every request in real time. No policy decision creates a blind spot in protection.