Identity Network for Synthetic Identity Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for verifying identities in network communications fail to effectively detect synthetic identities, which are combinations of real and fictitious attributes, making it difficult for businesses to distinguish between real and fraudulent identities, especially in online interactions where the longevity of synthetic identities can make them appear legitimate.
Innovation Solution
A system and method that utilizes an identity network to generate an assessment metric based on a device behavior score, user profile score, and fraud profile score, which aggregates data from device interactions and historical applications to determine the likelihood of an identity being synthetic, allowing relying parties to assess and manage risk or require further verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional identity verification methods are used, then verification process is simple, but synthetic identities cannot be effectively detected
Solution Approach 1:
The verification system is divided into multiple independent scoring components: device behavior score (analyzing interaction patterns), user profile score (evaluating demographic consistency), and fraud profile score (checking against known fraud indicators). Each component independently evaluates specific aspects of identity authenticity, and their results are aggregated to form a comprehensive assessment metric, enabling detection of synthetic identities without requiring a single complex verification system
Solution Approach 2:
An identity network acts as an intermediary between relying parties and users, receiving verification requests, computing assessment metrics based on multiple data sources, and returning results. This intermediary layer handles the complexity of multi-factor analysis internally while presenting a simple verification interface to external parties, resolving the contradiction between detection accuracy and system complexity
2Measurement precision
If multiple verification checks are implemented, then synthetic identity detection improves, but processing time increases
Solution Approach 1:
The system pre-computes and stores baseline profiles for devices, users, and fraud patterns before verification is needed. During actual verification, the system retrieves these pre-prepared profiles and compares them against current submission data, rather than performing complete analysis from scratch. This preliminary preparation significantly reduces real-time processing time while maintaining multi-factor detection accuracy
Solution Approach 2:
The verification system implements progressive evaluation, skipping detailed analysis of certain factors when initial checks indicate high confidence in authenticity. For example, if device behavior patterns strongly match legitimate users and fraud profile checks return clean results, the system can expedite the process by reducing the depth of user profile analysis, thereby maintaining accuracy while reducing processing time for low-risk cases
3Measurement precision
If comprehensive data analysis is performed, then assessment accuracy improves, but computational resources increase
Solution Approach 1:
The system performs partial analysis by focusing computational resources on the most discriminating factors for each verification case. Rather than uniformly analyzing all available data, the system identifies and prioritizes key indicators (such as device behavior anomalies or fraud profile matches) that provide the highest information value, performing detailed analysis only on these critical aspects while using simplified evaluation for other factors, thereby maintaining accuracy while reducing overall computational burden
Data Source
AI summary
Systems and methods are provided for use in identifying synthetic identities. One example method includes receiving a request from a relying party for an identity asserted by a user to the relying party, where the request includes identity data indicative of the identity, feature data associated with the user asserting the identity, and a device ID for a communication device of the user. The method also includes calculating a fraud profile score, based on the identity data and a data structure of known fraud profiles, and aggregating the fraud profile score and at least one of a device behavior score, a user profile score, and/or an exposure behavior score into a metric indicative of a likelihood that the identity asserted by the user is a synthetic identity. The method then includes transmitting the metric to the relying party.