
- The Department for Work and Pensions is trialling a machine learning model to help select Universal Credit claimants for Targeted Case Reviews.
- DWP permanent secretary Dame Sarah Healey stated that human consideration remains central and the AI will not make final decisions on benefit entitlements or suspensions.
- The anti-fraud initiative aims to save £13bn by the end of the decade, but research shows only 21 per cent of reviews currently uncover payment errors.
- Experts from the Administrative Fairness Lab warned of potential bias, noting a previous DWP tool disproportionately flagged older, disabled, and non-UK claimants.
- The DWP confirmed that specific criteria used to flag cases will remain confidential to protect the integrity of the welfare system.
IN FULL


.jpeg?width=1200&auto=webp&trim=1%2C0%2C0%2C0)