Skip to main content
The Users API lets you query the risk profile that AntifraudAI has built for any user in your system. As you submit transactions and events for a user over time, AntifraudAI builds a continuously updated risk profile that aggregates signals, score trends, and behavioral history. You can read this profile at any time to inform your own application logic — for example, to display a trust badge, adjust withdrawal limits, or trigger an additional verification step.

​
GET /users//risk-profile

Returns the current risk profile for a user, including their aggregated risk score, risk trend direction, active signals, and a summary of their screened event history.
​
string
required
Your internal identifier for the user, exactly as passed in the user_id field of transaction or event submissions.
cURL

​
Response Fields

​
string
Your internal user identifier, echoed back from the path parameter.
​
integer
The user’s current aggregated risk score from 0 (lowest risk) to 100 (highest risk). This score is recalculated with each new event and incorporates behavioral patterns across the user’s full history.
​
string
The direction the user’s risk score has been moving over the past 30 days. One of:
  • increasing — risk is trending upward
  • stable — risk is not meaningfully changing
  • decreasing — risk is trending downward
​
array of strings
The set of risk signals currently active on this user’s profile, e.g. ["multiple_devices", "high_velocity_login"]. Signals are automatically cleared when the underlying behavior normalizes.
​
integer
The lifetime count of transactions and events that have been submitted for this user. Includes all event types.
​
string
ISO 8601 UTC timestamp of the most recent transaction or event screening for this user.
Example response:
Risk profiles are only available for users who have had at least one event or transaction screened through AntifraudAI. Querying the profile for an unknown user_id returns a 404 not_found error.
Use the risk_trend field alongside overall_risk_score for a richer signal. A user with a score of 45 but an increasing trend may warrant closer attention than a user with a score of 55 but a decreasing trend.