Sofia: Officials who manage cohesion money in Bulgaria spent August reading a document that quietly changes how Brussels will grade them. On 28 July the European Commission published its 2025 report on the protection of the Union’s financial interests, and for the first time the text refuses to treat detection figures as the headline. It grades the entire chain instead, from prevention through investigation to the moment cash actually returns to the budget.
The headline numbers repay a careful reading. Member States and EU bodies logged 13,010 irregularities during 2025, worth roughly 2.1 billion euros. Of those, 986 carried a fraud classification and involved 274.3 million euros. The count of reported irregularities fell by 7.6 percent against 2024, yet the money attached to them climbed by 12.8 percent. Fewer cases, bigger cases: that pattern points at organised schemes rather than sloppy paperwork, and it explains why the Commission keeps pushing risk-scoring tools such as Arachne+ toward managing authorities that still resist them.
Recovery is where the report sounds most confident. Across 2016 to 2025, where follow-up finished, national and EU authorities recovered 96 percent of the sums OLAF recommended for recovery. Revenue cases account for the bulk of that, with 4.53 billion euros collected against 4.18 billion recommended, because customs debts attach to identifiable importers and national customs services chase them hard. Expenditure tells a thinner story. Indirect management recovered 41.1 million euros, and only 42 percent of the recommended amount was ever formally established as a debt in the first place. A high recovery rate calculated on a small established base flatters the system more than it should.
The forward-looking half of the report matters more for capitals. The Commission wants mandatory national anti-fraud strategies in place before the next multiannual financial framework starts, alongside firmer reporting duties and a review of the legal architecture governing OLAF, the European Public Prosecutor’s Office and the Commission’s own controllers. A Communication due later this year will set the shape of that review, and legislative proposals could follow.
National administrations will push back on two fronts. Several finance ministries argue that they already run domestic strategies and that a Brussels template adds reporting work without adding arrests. Others note that the report itself concedes the weak link sits after a file reaches a courtroom, where prosecutors share little with administrative authorities and recovery stalls for years. A mandatory strategy document does not fix a prosecutor who stops answering emails.
Prevention offers the cleanest evidence for the Commission’s case. Over the past decade OLAF and Commission services together stopped more than 658 million euros from being spent improperly, mostly through exclusion mechanisms, early detection and better information sharing between paying agencies. That figure never appears in a recovery table because the money never left. Piotr Serafin, the Commissioner responsible for the budget and anti-fraud, framed the report as a step toward transparency rather than a finished job, which is an unusually modest claim for a July press release.
The recurring fraud pattern has not changed much. Falsified invoices, invented eligibility criteria and misdeclared values still dominate the fraudulent files, while misdeclaration of goods at customs and breaches of public procurement rules lead the non-fraudulent ones. Those are administrative failures as much as criminal ones, and they will not yield to a new reporting indicator alone.
What the 2025 report does achieve is a shared scoreboard. For the first time the Commission, Member States, OLAF and the EPPO report into one framework, which means next year’s edition can show whether the anti-fraud strategies argument was worth having.
