Should AI be judging our credit worthiness?
Summary
Identity fraud sharply lowered Sophie Hope’s credit score and led a lender to deny her mortgage application. AI-based lending tools may help assess people with limited credit histories, but experts warn that these systems can repeat past inequalities and that consumers need ways to understand and challenge decisions.Key Facts
- Fraudsters opened catalogue accounts in Sophie Hope’s name and bought items including a PlayStation 4 and Nike socks.
- Debt linked to the fraudulent accounts accumulated for three to four years before Hope discovered it while applying for a mortgage.
- UK credit scores are produced by Experian, Equifax and TransUnion, using broadly similar data.
- Thin-file applicants, including young people and recent immigrants, may have limited credit histories for traditional scoring systems to assess.
- US firms Upstart and Zest AI offer lenders machine-learning tools that analyze additional information when assessing credit applicants.
- Upstart says its models use more than 2,500 variables to assess borrowers.
- Experts warn AI lending models could reproduce social inequalities reflected in past decisions, and say consumers need to understand and challenge decisions made about them.
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.