HM Revenue & Customs’ (HMRC) must strengthen governance and risk management around data-driven fraud and error initiatives after weaknesses in oversight led to significant disruption for Child Benefit claimants, according to a new report from the National Audit Office (NAO).
The watchdog found that HMRC’s use of Home Office travel data identified a way to tackle one of the largest causes of Child Benefit fraud and error, but changes made during the programme’s first rollout compromised delivery, negatively affected eligible claimants and damaged public trust.
The findings come as government departments increasingly look to data sharing, analytics and AI to improve services, reduce fraud and deliver efficiencies.
Child Benefit is paid to around 6.9 million families and supports approximately 11.7 million children. In 2024-25, HMRC estimated that fraud and error resulted in £270 million of overpayments, with one of the most common causes being claimants who had moved overseas without informing the department.
To address the problem, HMRC identified an opportunity to use international travel data held by the Home Office to investigate potential non-residency among Child Benefit claimants. A pilot carried out in 2024 matched airline data against Child Benefit records and was designed to identify people who may no longer have been eligible for support.
The pilot proved successful. HMRC estimated it prevented around £15 million in incorrect payments and projected that a wider rollout could generate gross savings of £366 million between 2025-26 and 2029-30, with a return on investment of around 14:1.
Rollout exposed governance weaknesses
However, the report found that key changes made when HMRC moved from the pilot to a full rollout in August 2025 significantly increased the likelihood that eligible claimants would be caught up in the intervention.
Most notably, HMRC removed an upfront check against its Pay As You Earn (PAYE) system that had previously been used to identify claimants with evidence of ongoing employment in the UK. During the pilot, this check removed 12 percent of potential matches before any enquiry was opened.
By October 2025, HMRC had opened enquiries into 23,794 claims and suspended Child Benefit payments while investigations were carried out. The NAO found that “many eligible claimants had been adversely affected by the intervention and had their Child Benefit payments suspended”.
The report notes that claimants reported financial and emotional impacts from the sudden loss of income, as well as difficulties proving their eligibility. It also states that there was “a negative impact on public trust, with widespread media coverage and Parliamentary scrutiny of the problems in the first rollout”.
As of April 2026, HMRC had confirmed 59 percent of the 23,794 suspended claims as eligible. During the first rollout, the department handled more than 22,500 calls from affected claimants.
If you liked this content…
Government not using data analytics effectively to tackle fraud
Collaboration in government isn’t a tech problem – it’s a human one
Blanket consultancy cuts risk false economy warns expert
GAD’s climate-reporting role spotlighted in major NAO review
People still want a human touch amid public sector’s AI push – research
No clear accountability
The report highlights significant shortcomings in governance arrangements during the transition from pilot to rollout.
According to the NAO, HMRC “did not appoint a single Senior Responsible Owner for the intervention or establish governance arrangements with clear accountability for approving major changes to the process”.
The watchdog found that key decisions were made without sufficient scrutiny or challenge and that there was “no due consideration of HMRC’s risk appetite or formal risk assessment of the implications of decisions”.
The report also concludes that HMRC’s focus on fraud detection and compliance meant it “did not sufficiently consider the impact of its intervention on affected claimants and their experience”.
In response, HMRC commissioned an internal audit, halted new enquiries and introduced a series of changes, including reinstating PAYE checks, simplifying evidence requirements and allowing claimants to confirm eligibility by telephone.
Lessons for digital government
The NAO is clear that departments should not be discouraged from using data to tackle fraud and error, but argues that innovation must be supported by stronger governance and oversight.
In its conclusions, the watchdog states: “HMRC’s pilot using Home Office travel data identified an innovative opportunity to reduce the impact of a major driver of Child Benefit fraud and error. However, weaknesses in governance and risk management meant that HMRC made changes for the first rollout of the intervention that compromised delivery and performance.”
Gareth Davies, head of the NAO, said: “It’s right that HMRC seeks new ways to tackle fraud and error. HMRC’s initial work on using travel data to investigate potential Child Benefit overpayments suggested it could secure significant savings for the taxpayer.
“However, missteps in implementing the first rollout meant HMRC did not strike the right balance between detecting fraud and error and managing the impacts on Child Benefit claimants. While HMRC should not be discouraged from pursuing innovative ways to reduce fraud and error, it must learn and apply the lessons for future initiatives.”
The NAO recommends that HMRC clearly articulates its risk appetite for future interventions, ensures performance metrics track customer experience alongside compliance outcomes, and designs processes that allow eligible claimants to prove their status as efficiently as possible. It also calls on the department to share lessons learned across government.
