Tenants declined by faulty TransUnion AI system

Occurred: January 2023

Can you improve this page?
Share your insights with us

US-based credit score company TransUnion was found to have wrongly reported criminal and landlord-tenant eviction records to third-parties based on a machine learning-powered automated system to detect 'higher-risk' property tenants.

Described as an 'enhanced analytics screening model', TransUnion Rental Screening Solutions' ResidentScore system was found in court to have falsely accused tenants of littering and other alleged misdemeanours due to inaccurate, misleading, or outdated data, resulting in tenants being unfairly declined when applying for rental properties, and losing their application costs.

In May 2023, 15 US state attorneys general urged (pdf) regulators to ensure that 'applicants for housing have access to all the data that is being used to make determinations of their tenant ‘worthiness''. Transunion agreed to a USD 11.5 million settlement for violating the US Fair Credit Reporting Act (FRCA) in September 2023.

The system, together with others like it, is also increasing costs and barriers to housing, according to a 2022 report from the US Consumer Financial Protection Bureau.

Operator: TransUnion Rental Screening Solutions
Developer: TransUnion Rental Screening Solutions

Country: USA

Sector: Business/professional services; Real estate sales/management

Purpose: Determine eviction likelihood

Technology: Risk assessment algorithm; Machine learning
Issue: Accuracy/reliability; Bias/discrimination - economic
Transparency: Governance; Black box

Page info
Type: Incident
Published: July 2023
Last updated: February 2024