Particle.news
Download on the App Store

UK Push to Curb Benefit Overpayments Advances as AI Flags Older Claimants at Far Higher Rates

New NAO findings tie DWP analytics to savings, highlighting sharp age disparities in referrals.

Overview

  • The National Audit Office reports overall benefit overpayments fell to 3.3% of spending in 2024–25 from 3.6%, with Universal Credit overpayments down to 9.7% from 12.4%.
  • DWP’s machine learning model, used since May 2022 to flag risky Universal Credit advance claims, is estimated to have saved about £4.4 million and is around three times more effective than baseline methods.
  • NAO and DWP fairness analysis shows claimants above state pension age were referred for investigation at nearly 50 times the rate of younger people, with older age groups and non‑UK nationals also over‑referred.
  • Targeted Case Reviews have scaled to about 6,000 staff, reviewing 1.15 million claims and generating an estimated £581 million in savings by March 2025.
  • Legislation moving through Parliament would enable targeted bank‑account checks and allow direct recovery of some overpayments, initially for Universal Credit, ESA and Pension Credit, with possible expansion and a government savings estimate of £1.5 billion over five years, as charities warn of privacy risks and the DWP points to safeguards and human oversight.