Asked by: Debbie Abrahams (Labour - Oldham East and Saddleworth)
Question to the Department for Work and Pensions:
To ask the Secretary of State for Work and Pensions, whether the (a) Integrated Risk and Intelligence Service, (b) Enhanced Checking Service, (c) Risk Review Team, (d) Enhanced Review Teams, (e) Universal Credit advances claims decision risk model, (f) Common Risk Engine, (g) General Matching Service, (h) Fraud Referral and Intervention Management System, (i) Targeted Case Review and (j) any other systems rely on artificial intelligence, machine learning or algorithmic processes for fraud detection.
Answered by Paul Maynard - Parliamentary Under-Secretary (Department for Work and Pensions)
As set out in the Department’s 2022/23 Annual Report and Accounts (page 308 to 310), DWP uses advanced analytics to tackle fraud and error. These analytics include a variety of sophisticated techniques including the use of machine learning to identify patterns in claims that could suggest fraud or error, so that these claims can be reviewed by relevant DWP teams such as the Enhanced Review Team. The final decision on benefit entitlement is made by a human caseworker.
The National Audit Office confirm that DWP have governance and processes in place to monitor the bias of these models.
Asked by: Debbie Abrahams (Labour - Oldham East and Saddleworth)
Question to the Department for Work and Pensions:
To ask the Secretary of State for Work and Pensions, whether any of the claim detection criteria in the targeted case review system are applied by or operated through artificial intelligence, machine learning or algorithmic processes.
Answered by Paul Maynard - Parliamentary Under-Secretary (Department for Work and Pensions)
As set out in the Department’s 2022/23 Annual Report and Accounts (page 308 to 310), DWP uses a range of advanced analytics to identify patterns in claims that could suggest fraud or error, so that these claims can be reviewed by relevant DWP teams including Targeted Case Review agents. The final decision on benefit entitlement is made by a human caseworker.
The National Audit Office confirm that DWP have governance and processes in place to monitor the bias of these models.