On May 1, 2026, the Netherlands Authority for Consumers and Markets (ACM) closed its abuse-of-dominance investigation into a large, internationally operating business-software company, stating bluntly that "the investigation has yielded insufficient evidence to establish a violation." The probe was not a minor inquiry: ACM opened it on September 30, 2025 with a dawn raid, suspecting the firm of charging "unreasonably high prices" for certain software and imposing "unfair conditions" on Dutch buyers who depend on it. Seven months later, the regulator walked away empty-handed.
That outcome will read, in some quarters, as a defeat for enforcement. It should read as the opposite: a competition authority applying an evidentiary standard and declining to manufacture a case it could not prove.
The case for the probe was real
Start with the strongest version of ACM's original suspicion, because it is genuinely strong. Enterprise software — ERP systems, databases, productivity suites — is the connective tissue of modern business, and once an organisation has built its operations around a vendor, switching is slow, costly, and risky. That dependence can hand a dominant supplier the leverage to raise prices well above competitive levels or to attach contractual terms a customer would never accept in a contestable market. Under Article 102 TFEU and the Dutch Competition Act, firms with market power carry what ACM itself calls a "special responsibility" not to exploit it. A dawn raid on a company suspected of doing exactly that is a proportionate investigative step, not an overreach. Captive buyers cannot vote with their feet, so someone has to check.
The problem is not the suspicion. The problem is proving it — and that is where the case collapsed.
Excessive pricing is the hardest case in competition law
Exploitative abuse through "unfair" or "excessive" pricing is notoriously difficult to establish, and deliberately so. Ever since the European Court of Justice's United Brands judgment in 1978, authorities pursuing an excessive-pricing theory have had to show both that a price bears no reasonable relation to economic value and that it is unfair in itself or against competing products. That demands a defensible benchmark for what the "right" price is — a counterfactual that, for differentiated, R&D-heavy software with no obvious comparator, is close to impossible to construct without the regulator effectively setting the price itself.
This is why excessive-pricing cases are rare across Europe and why ACM was right to drop one it could not substantiate. A competition authority that brings exploitation cases on thin evidence does not protect buyers; it injects uncertainty into every pricing and contracting decision a successful firm makes, and it risks punishing the returns that justify the investment in the product in the first place. Restraint here is not timidity. It is the rule of law: no infringement finding without proof.
Contrast with a case ACM actually won
ACM is not a passive regulator, and the software closure should not be mistaken for one. The decisive contrast is its long-running fight with Apple. In August 2021, ACM found that Apple had abused its dominant position toward dating-app developers by forcing them onto its in-app payment system, barring references to cheaper external options, and charging a 30 percent commission (15 percent for small developers). On June 16, 2025, the Rotterdam District Court upheld that decision, confirmed €50 million in penalties Apple had already paid, and warned of a further €5 million per week if Apple did not fix its commission terms.
The Apple case succeeded because the theory of harm was concrete and observable: specific contractual restrictions on developers' freedom to transact, behaviour ACM and a court could examine without inventing a fair-price formula. The software case failed because exploitation through "high prices" offered no such tractable hook. The lesson for proportionate enforcement is clear — go after conduct you can define and remedy, not abstract claims about whether a price is too high.
The risk is what comes next
There is a less reassuring backdrop. On May 27, 2026, the Dutch government blocked Kyndryl's roughly €100 million acquisition of cloud provider Solvinity on public-interest grounds — the first time the country's Investment Screening Bureau had stopped a US acquisition since 2020. The minister conceded she could not explain why the deal threatened the public interest, even while praising the contribution of American technology firms. Kyndryl called the decision a "politicisation" of the deal.
The worry is that when antitrust tools fail to deliver — as they just did against the software vendor — the temptation grows to reach for blunter instruments dressed as "digital sovereignty," which carry no comparable evidentiary discipline. An abuse case demands proof before a court. A foreign-investment veto demands only a ministerial signature. If the policy goal is to discipline powerful technology suppliers, the disciplined path is the one ACM just demonstrated: investigate hard, charge only what you can prove, and remedy specific conduct. The closure of the software probe is not a regulatory failure. It is what a credible, evidence-based competition regime looks like — and a standard the rest of Dutch tech policy should be held to.