DWP utilizing new AI system for Universal Credit and PIP claims – here is the way it works

The Department for Work and Pensions has revealed new details of how it uses artificial intelligence (AI) to identify overpayments and potentially fraudulent claims across a range of benefits.

The Department for Work and Pensions (DWP) is cracking down on benefit fraud and overpayments using cutting-edge AI technology.

The AI tool dubbed ‘intelligent claims analysis’ has been deployed to sniff out inconsistencies in Universal Credit and Personal Independence Payment (PIP) claims, according to the government’s official algorithm database.

The AI cross-checks new applications against previous data and even looks at external databases like tax and employment details. By reviewing key indicators and past data patterns, the system dishes out a “risk score”, boasting an 89% success rate at picking claims that need closer inspection.

But it won’t be AI making the final decisions. A human agent will step in to verify any dodgy-looking cases before any drastic measures are taken.

Once the alarm bells ring, the case is registered in the department’s fraud and compliance system for a thorough manual review by a specialist. This layer of human judgement ensures that benefits aren’t chopped without good reason, reports the Express.

This technology has been being used since mid-2021, but the DWP only publicly shared the details this week.

The department said that the system has crunched through over 1.2 million cases, shaving 60,000 hours by cutting back on manual checks.

The early models, used from 2021 to 2023, did churn out a higher than expected rates of false positives and getting it wrong by flagging legitimate claims.

Director at Welfare Rights UK James Walker said: “While ensuring taxpayers’ money is used appropriately is important, we must also ensure that AI-driven fraud detection does not unfairly target those who genuinely need support. The DWP must provide clear explanations and appeal routes for those affected by AI-flagged claims.”

This news surfaces as Keir Starmer amps up Labour’s drive to supercharge AI use in public services, with the PM touting AI’s potential to boost efficiency across welfare, healthcare, and education sectors.

Yet, the fraud detection tool remains one of the few AI systems the DWP has revealed on the algorithm transparency register, a mandate for government departments for over a year. Despite having an internal list of AI tools, the DWP has declined to publish it fully, citing the need for discretion in information release for effective government operation, especially when responding to a Freedom of Information Act request in September: “Public authorities like the DWP must retain control over how and when they release information. The ability to manage disclosures is essential to the effective operation of government functions.”

The measures are expected to be primarily aimed at Universal Credit, which has a significant number of claimants, with 6.4 million as of January 2024. Officials might also focus on those receiving Jobseeker’s Allowance, Employment and Support Allowance, and Housing Benefit, as these benefits have traditionally seen higher levels of irregularities.

For the latest breaking news and stories from across the globe from the Daily Star, sign up for our newsletters.