Brussels: When the European Anti-Fraud Office reported in its 2025 annual review that it had recommended the recovery of almost 600 million euro in misused EU funds, the figure landed as both a measure of enforcement and a reminder of how much leaks from the system in the first place. OLAF closed 209 investigations over the year and opened 254 new ones, recommending the clawback of some 597 million euro while preventing a further 18 million from being spent improperly. Read one way, that is a watchdog doing its job. Read another, it is a snapshot of the scale of irregularity that a single year of EU spending still generates.
The composition of the caseload is as telling as the totals. OLAF’s 2025 work spanned complex financial irregularities, cross-border illicit trade, customs violations and environmental fraud, with recurring themes of conflict of interest, procurement manipulation and inflated costs. These are not exotic crimes; they are the predictable failure modes of any large funding architecture where money flows from a central budget through national and regional intermediaries to thousands of final beneficiaries. Each handover is a seam, and seams are where fraud finds purchase.
Two strands of the year’s work carry particular weight. The first is sanctions enforcement. OLAF continued to pursue the circumvention of EU measures against Russia and Belarus, including a coordinated cross-border investigation into the suspected illicit export of more than 760 transport vehicles, and it provided support to Ukrainian customs and law-enforcement authorities. This places the anti-fraud office squarely inside the geopolitical machinery of the sanctions regime, a role it did not occupy a decade ago. The second strand is the misuse of agricultural and rural development funds, illustrated by a Hungarian case in which OLAF’s findings prevented roughly half a million euro from being drawn down improperly. Cohesion and farm spending remain the categories most exposed precisely because they are the largest and the most decentralised.
The longer record lends perspective. Over the past decade OLAF investigations have led to 6.8 billion euro being recovered to the EU budget and 873 million euro being prevented from improper spending. Those are substantial sums, yet they invite a harder question that the headline recovery figure tends to obscure: recovery is recommended, not guaranteed. OLAF can investigate and advise, but actual collection depends on national authorities and judicial follow-through, and the gap between a recommendation issued and money returned is where much of the system’s credibility is won or lost.
That gap is also why the institutional context matters. The European Court of Auditors devoted a special report in 2025 to the bodies fighting EU fraud, and the existence of the European Public Prosecutor’s Office now gives some cases a criminal route that OLAF’s administrative recommendations previously lacked. The emerging architecture is one of overlapping mandates, OLAF for administrative investigation, the EPPO for prosecution, national agencies for execution, which improves coverage but raises coordination costs. The risk is not that fraud goes unexamined but that responsibility becomes diffuse enough for recoveries to stall between institutions.
There is a tempting but misleading reading of these numbers, which treats a higher recovery figure as straightforwardly good news. It is better understood as ambiguous. A larger recovery total can reflect sharper detection, but it can equally reflect more fraud to detect, and the report alone does not let an outside observer cleanly separate the two. The more defensible conclusion is that EU spending remains structurally exposed at its disbursement points, and that the size of the annual recovery is less a victory lap than a running estimate of that exposure.
For a Union preparing its next multiannual budget amid pressure to spend faster on defence, competitiveness and reconstruction, the OLAF report is a quiet caution. Speed and simplification are reasonable goals, but each removed control is also a removed safeguard. The 597 million euro recommended for recovery in 2025 is the price tag on that tension, and it will only grow more visible as the money does.




