By Matt Horne, Director of Intelligence and Investigations, Clue Software
The Public Authorities (Fraud, Error and Recovery) Act 2025 hands the PSFA significant new powers – to investigate, levy civil penalties and recover money on behalf of public authorities. It is the most significant escalation in public sector counter-fraud enforcement in years.
Importantly, several provisions go further than enforcement after the fact. The DWP-specific elements of the Act are proactive – giving the department access to real-time data to detect breaches by benefit recipients before fraud becomes entrenched. New bank account monitoring powers will allow DWP to identify claimants with undeclared savings above eligibility thresholds or additional income streams that have not been declared: two of the most persistent and costly forms of benefit fraud.
But legislation alone does not solve the underlying problem. The harder question – the one senior leaders across departments and arm’s-length bodies need to be asking now – is whether their organisations are structurally equipped to detect fraud early enough for any of those powers to matter. For many, the honest answer is not yet. But the tools, frameworks and political will to change that are now in place.
The gap between standard and reality
The Government Functional Standard GovS 013 sets out what adequate counter-fraud controls look like – and the PSFA audits departments against it. The Cross-Government Fraud Landscape Report, published in 2026 using 2023/24 data, is instructive: DBT, health and education bodies detected and prevented the most fraud. Beyond that, the picture is varied. Many organisations simply do not have the capacity to detect fraud, let alone respond to it.
This is not a question of willingness. The professionals responsible for counter-fraud work are skilled and acutely aware of the standards expected of them. The problem is that they are too often held back by technology debt, legacy systems and slow organisational change. One of the most significant shifts I have observed is fraud teams beginning to adopt the intelligence-led prioritisation tactics long used in law enforcement. That shift is the right one – and the new legislative framework creates real momentum for organisations willing to build on it.
The early warning problem
Where does early-warning intelligence fail most consistently? In my experience, the answer is all three: technology, process and culture – but the root cause is usually capacity. Counter-fraud professionals are overwhelmed not by a lack of data, but by too much of it. Without the ability to triangulate signals with risk and intelligence insights, without clear decision-making frameworks, and without structured case management and consistent data sharing between departments, organisations drown in noise rather than acting on it.
The solution is not more technology. It is designing the response process end-to-end before embedding data analytics – so that when a signal is identified, there is a clear, resourced pathway from detection to action. This is an organisational design
challenge as much as a technology one – and it is entirely within the reach of bodies that choose to prioritise it.
A threat environment that is not standing still
The threat facing public services is not static. AI-enabled fraud, synthetic identities and deepfake impersonation are operational realities. AI-enabled attacks make welfare fraud possible remotely, at speed and from multiple jurisdictions simultaneously. Research from LexisNexis has identified millions of synthetic “Frankenstein” identities posing multi-billion-pound fraud risks to UK organisations – and public services are not immune.
More broadly, fraud is no longer a standalone threat category. It is increasingly cyber-enabled, cross-border and connected to wider criminal networks – drugs, money laundering and human trafficking. The investigative model that worked a decade ago is no longer fit for purpose, and organisations that invest in building better capability now will be significantly better placed than those that wait.
What intelligence-led actually means in practice
Moving from reactive investigation to a structured, intelligence-led operating model is not a technology project. It is an organisational one – and it starts with clarity of purpose: proactive identification of fraud networks, predictive targeting, disruption rather than prosecution. In the public sector, this aligns to the National Intelligence Model – a framework that provides genuine structure for how intelligence is gathered, assessed and acted upon.
From there, the practical requirements follow: data analytics capability, robust reporting tools, structured case management, and the training to use them effectively. Organisations need to agree the decision points that will be intelligence-driven – case selection criteria, resource allocation, intervention types – and set outcome-based metrics that go beyond case counts to measure fraud prevented and risk reduced. None of this requires unnecessary bureaucracy. It requires intentional design.
With new powers comes responsibility
The PSFA’s new enforcement unit – providing investigative services on behalf of departments and arm’s-length bodies that lack the capability to respond alone – is a significant and welcome development. Many public sector organisations are established to deliver policy missions, not investigate fraud and recover debt. Having a central resource to call on fills a genuine gap.
But with expanded powers comes expanded responsibility. In the wake of the Post Office Horizon scandal, it is both right and necessary that the use of these powers – particularly the DWP’s real-time data access provisions and the PSFA’s enforcement remit – is subject to a strong, robust assurance model. Public trust in the state’s ability to exercise investigative powers fairly has to be earned and maintained, not assumed. Getting the oversight framework right is not a bureaucratic consideration – it is a fundamental condition of the Act delivering the outcomes it promises.
The enforcement unit should be understood as a floor, not a ceiling. It cannot replace the investment in detection, intelligence and structured investigation capability that every public body needs to build for itself.
Seizing the moment
The Act has raised the bar – and for organisations willing to act, it represents a genuine opportunity to build the resilient, intelligence-led counter-fraud capability the current threat environment demands. The legislative framework is in place. The PSFA has the tools. The question now is whether departments and arm’s-length bodies will use this moment to close the gap between the standard expected of them and the reality of how they currently operate.
The professionals in the room already know what needs to happen. The question is whether their organisations will give them the tools – and the mandate – to make it happen.
Matt Horne is Director of Intelligence and Investigations at Clue Software, the intelligence and investigation platform trusted by the Home Office and NHS Counter Fraud. Clue serves government, law enforcement, financial services and corporate clients worldwide.
Recent Comments