Technology news and discussion on TalkFuse.
Should AI be judging our credit worthiness?
Sophie Hope’s mortgage application was rejected after identity fraud damaged her credit score, despite her previously excellent rating. Her case highlights how heavily access to borrowing and services can depend on credit records that may be inaccurate or fail to reflect a person’s actual ability to repay.
UK credit scores from Experian, Equifax and TransUnion are based largely on past financial data, which can disadvantage people with little credit history. In the US, AI lenders such as Upstart and Zest AI use machine-learning models that consider more than 2,500 variables, potentially helping lenders assess affordability and include more borrowers. Experts warn that AI may reproduce past social inequalities, so people need to understand and challenge decisions made about them.
In short
- Identity fraud left Sophie Hope with a very poor credit score.
- AI could help assess borrowers with limited credit histories.
- Experts warn automated lending must be accountable and fair.