Baroness Sherlock (Lab)

My Lords, I regard that as a challenge. I am confident that I can assure the noble Viscount in the way that he wants to be. As I have said repeatedly—ad nauseum, to be fair—throughout Committee, the Government have a responsibility to tackle fraud and error and ensure that they are minimised. Fraud and error in the social security system were responsible for the overpayment of almost £10 billion in 2023-24. We recognise that there are opportunities for technology and data to help to identify potential fraud and error risks while also understanding the need to ensure their safe and effective use. I remind the Committee that, while the DWP is improving its access to relevant data through this Bill, we are not introducing any new automated decision-making measures in the Bill. I will explain why this amendment is unnecessary, but I will pause briefly and digress. The noble Baroness, Lady Bennett, was commendably brief in her digression, and I will be commendably brief in mine. The Committee has at different points queried the role of automated decision-making, so I will put this point on the record. I start with the eligibility verification measure, a data-requiring measure to help the DWP identify where claimants do not meet the eligibility criteria for the benefit they are receiving. The DWP will review all information received, and DWP staff will make any decisions about entitlement where potential fraud or error is identified. No decisions will be taken using EVM data alone. Decisions about entitlement will be made only once the DWP has made further inquiries. Similarly, as previously debated, there will be no automated decision-making from the information obtained under the PSFA’s or the DWP’s information-gathering powers when we are investigating specific cases of suspected fraud. Again, decisions on the use of the new debt recovery powers will always be made by a trained member of staff.