On Monday 2 November 2026, Pro MP Barbara Kathmann submitted a private member's bill to the Tweede Kamer. It would classify manipulative app design, with the infinite scroll as the headline example, as an unfair commercial practice under Dutch consumer law (DutchNews). The reported scope covers social media, games, dating apps and chatbots. Euronews adds streaming, autoplay and push notifications, and describes a "fairness by design" duty of care under which platforms would have to show their services are not addictive (Euronews). Two caveats apply. This is not a government bill. And the press reports do not say who would enforce it, what penalties it carries, or how a platform would prove a negative. We have not seen the bill text, so those points stay open.
The strongest case for the bill
The argument for acting deserves a fair statement. Kathmann says existing rules are "geared to physical products" while platforms have become "the new high street or marketplace" (DutchNews). Consumer law has long barred practices that distort people's economic decisions. If a product is built to keep users online and push them into spending money, it is reasonable to ask why the interface should be exempt. Design choices are also not speech in the ordinary sense: no one's opinion is suppressed when a feed stops auto-refilling. And the pressure is not only Dutch. The article reports that Kathmann is working with MPs in Germany, France, Belgium, Luxembourg, Austria, Italy and Portugal on similar laws. She says she would prefer EU-level action but that it "would take too long."
Why this is an intermediary liability story
The Digital Services Act (Regulation 2022/2065) built EU intermediary liability around content. Under Article 6, a hosting provider is not liable for information it stores unless it has actual knowledge of illegal content and fails to act expeditiously (DSA, EUR-Lex). That model asks what a user posted and what the platform knew about it.
The Dutch bill asks a different question: what did the platform build? Liability attaches to the operator's own conduct, whatever any user says. That makes it a quiet but real shift in how responsibility is allocated. It also explains the appeal. Notice-and-action regimes force platforms into judgments about lawful speech, which is where over-removal and free-expression risk concentrate. Design rules avoid that, at least in principle.
Where the proportionality problem sits
The trouble is definitional. "Infinite scroll" is a concrete pattern. "Manipulative algorithms" and "engagement pressure" are not. A ranking system that surfaces what a user is likely to enjoy is, from another angle, a recommender doing its job. If the legal test is whether a service is "addictive," the statute hands regulators and courts a standard that is hard to measure and easy to stretch.
The feed is also where speech and design overlap. How content is ordered, paginated and recommended shapes what people read and share. A broad ban on algorithmic curation risks pushing platforms toward chronological or heavily moderated feeds that are cheaper to defend legally, and that is a speech effect even if no content is removed. The reported duty to prove a service is not addictive compounds this. Reversing the burden of proof means a small Dutch games studio or a startup chatbot carries the same evidentiary load as a large platform with a compliance department.
The EU layer
The bill also sits awkwardly in an already crowded framework. Article 25 of the DSA bars online platforms from deceptive interface design, but it does not apply to practices already covered by the Unfair Commercial Practices Directive 2005/29/EC or the GDPR (DSA, Article 25(2)). Routing manipulative design through consumer law, as the Dutch bill does, therefore lands it in the UCPD, which is a harmonising directive. National legislators have limited room to add stricter rules of their own within its scope. That raises a real question about whether a national design ban can survive alongside it.
The stakes are not trivial. As amended by the 2019 Omnibus Directive, the UCPD lets national authorities impose fines of at least 4% of annual turnover in the Member States concerned for widespread infringements (UCPD, EUR-Lex). Euronews reports that the Commission is due to table a Digital Fairness Act this November, with an EU-wide addictive design measure among its key elements. We could not confirm from official Commission pages that the proposal has been published. If it is, a Dutch statute may be overtaken, or at least pre-empted, within months.
The Dutch Parliament has already asked for an EU-level answer. A motion dated 20 February 2025 (Kamerstuk 26643, no. 1302, by MP Kahraman and others) called for a European ban on addictive and polarising designs on social media (Tweede Kamer motion). A national bill and a Commission proposal running in parallel could leave operators facing two overlapping definitions of the same practice.
What a proportionate version looks like
The sensible path keeps the principle and narrows the mechanics:
- Name the practices. List specific patterns, such as countdown pressure, hidden cancellation flows or autoplay defaults for minors, in a blacklist, as the UCPD already does for other practices. Do not rely on an open-ended "addictive" test.
- Keep the burden on the regulator. Require evidence of consumer harm or distorted decisions, not a duty on every operator to prove a negative.
- Protect user choice. Make non-algorithmic or chronological feeds an available option rather than mandating one design, which preserves user control without dictating product architecture.
- Wait for Brussels. If the Digital Fairness Act text appears, align with it rather than build a parallel regime.
Bottom line
The bill is an honest attempt to regulate platform conduct without regulating speech. That is a better instinct than another content-removal mandate. But a private member's bill built on an undefined "manipulative" standard, and on a burden of proof that falls on operators, is likely to be vague enough to chill product experimentation and uncertain enough to collide with EU harmonisation. The design-liability idea is worth pursuing. It needs the precision that consumer law, at its best, already demands.