State pensioners could be at risk of “serious harm” from the increased use of artificial intelligence (AI) in DWP services, with concerns mounting over its implementation. The Department for Work and Pensions (DWP) is looking to expand AI usage to streamline Jobcentres and support services, which it claims will improve the system for the most vulnerable, including state pensioners.
However, this week a charity warned that “rushed implementation” could lead to problems, with “marginalised groups” potentially facing bias and discrimination if historical data is used. Shelley Hopkinson, head of policy and influencing at Turn2us, set out her concerns, saying: “AI has the potential to improve speed and consistency in the social security system, but errors from rushed implementation could cause serious harm to people relying on support.
“We are concerned that the historical data used to train AI may perpetuate bias and discrimination against marginalised groups.” The DWP’s machine-learning programme, which is designed to detect Universal Credit fraud, was found to disproportionately target certain demographics, as revealed by an investigation by The Guardian. It highlighted biases related to age, disability, marital status, and nationality.
Further research indicates that around 200,000 individuals were mistakenly probed for housing benefit fraud due to flawed algorithmic decisions. Ms Hopkinson emphasised to The Big Issue the importance of a careful approach to AI integration, stating: “AI integration should be guided by consultation and transparency so that the system meets the needs of everyone and builds trust that decisions are fair and accurate.
“This must come with clear accountability and safeguards to allow for challenge. AI must work for people, not against them, prioritising people’s lives and wellbeing in decision-making.” Meanwhile, research from the Tony Blair Institute has suggested that the DWP could free up some 40% of its time using AI tools, which could translate into nearly £1billion in productivity gains each year.
Sir Peter Schofield, the DWP’s permanent secretary, highlighted several ongoing initiatives where AI has already been beneficial, including the Health Transformation Programme and the Service Modernisation Programme. Prime Minister Keir Starmer recently hailed AI as a catalyst for improvements in both public services and private sectors, asserting that it will “make our public services better” and “boost growth”.
Several experts have also recently raised concerns about plans for new powers for DWP officials to request banking information, in cases of suspected fraud. The new measures will come in as part of the upcoming Fraud, Error and Debt Bill.
Ben Fleming, financial crime analyst at Ocean Finance, warned that innocent claimants could be wrongly targeted. He said: “Mistakes do happen, and the stress of being investigated can be overwhelming.”
He urged the DWP to look at other measures to prevent benefit fraud, suggesting: “Improving ID checks or cross-referencing information with other systems could help identify false claims before payments are made. However, this should be done carefully so it doesn’t create extra barriers for people who genuinely need help.”