This week, Brussels unveiled its long-awaited Digital Fairness Act (DFA), paired with measures to reinforce cross-border consumer protection and streamline existing safeguards. The package could reshape how the EU polices commercial conduct and curb business models that undermine fundamental rights—a substantial undertaking, yet one the Commission has pursued seriously for years. The impact assessment projects the DFA could avert €13.7 billion in annual consumer harm, accounting for lost time and other costs.

The DFA would supplant the Unfair Commercial Practices Directive with a directly applicable Regulation. Its centerpiece: fresh bans on manipulative interface design, addictive features, and problematic practices spanning influencer marketing, gaming, subscriptions, personalized pricing and AI agents making decisions on users' behalf.

Some manipulation will finally be banned

Among the DFA's strongest elements is a blacklist of ten prohibited digital deceptive design tactics—the so-called "dark patterns": bogus alerts, repeated nudges to reverse choices, deliberately opaque cancellation pathways and comparable stratagems. While welcome, this move highlights a puzzle: if regulators can outlaw certain manipulative designs outright, why does the EU hesitate to apply the same logic to addictive features that companies deliberately engineer?

Article 17 acknowledges that autoplay, infinite scrolling, engagement streaks and certain AI interactions drive compulsive overuse. For consumer law, recognizing that continuous service use counts as a commercial decision marks progress. Yet the proposal's remedy is largely an off switch. Commission documents reveal officials weighed requiring attention-maximizing features to be disabled by default. That option was shelved partly over worries about disrupting business models and user experience—reasoning that warrants scrutiny.

People plainly enjoy engaging platforms and tailored suggestions. But enjoyment says little about whether firms should constantly test how to extend our time online beyond what we intend. The Commission's own consultation data complicates the case: among respondents backing EU action on addictive design, 58% preferred allowing consumers to opt in rather than facing automatic exposure.

History offers a cautionary tale. Cookie consent, once a valid legal safeguard, became an obstacle course that frustrated users—ammunition for those arguing against EU privacy rules. Layering more controls, settings and choices risks repeating that cycle: frustration will mount, industry will denounce Brussels for spoiling experience, and calls to weaken rules (rebranded as "simplification") will follow. A legal gap also looms: could addictive features already violate prohibitions on manipulative design and unfair practices even with an opt-out available? That relationship requires clarification, since an off switch should never excuse manipulation.

Apparently, being vulnerable is no longer enough

Consumer law has long distinguished between the average consumer and those facing particular vulnerability, allowing authorities to judge unfairness from the perspective of vulnerable groups. Yet digital systems have made that line harder to hold. Vulnerability online is situational: financial strain, fatigue, emotional turmoil or isolation shift how people respond to commercial pressure. Firms track behavior, infer weaknesses and tailor tactics accordingly. Even ordinarily prudent individuals fall prey to systems optimized to exploit predictable responses.

The Commission's impact assessment concedes these limits. But between the leaked draft and final text, language shifted: the draft covered consumers vulnerable due to disability, age, credulity or personal circumstances; the official version narrows this to a specific misfortune or circumstance grave enough to impair judgment. That tightening could make it harder to assess practices from the angle of those facing temporary or situational vulnerabilities.

The general unfairness ban still applies universally. Yet it is puzzling why the Commission chose a more restrictive framing when its own research documents how digital firms exploit behavioral weaknesses. The EU should preserve stronger safeguards for those facing particular vulnerabilities—while acknowledging these can be fleeting, affect anyone and sometimes stem from commercial practices themselves. That demands companies identify and forestall foreseeable exploitation, rather than asking people to prove their circumstances warrant protection.

Personalized practices sharpen the dilemma. The proposal restricts certain harmful personalized pricing when traders know users are minors or particularly vulnerable. Yet firms increasingly detect and exploit vulnerabilities that may be temporary, hard to categorize, or deliberately manufactured by the commercial setting. Why should consumer law set such a high bar for vulnerability when identifying and leveraging ordinary weaknesses has become a commercial competency? Perhaps the focus should shift: less time asking whether people are vulnerable enough to merit protection, more time stopping businesses from creating or exploiting those vulnerabilities.

'Fairness by design,' but who's responsible for proving it?

The Commission's impact assessment explored a general fairness-by-design duty, backed by many civil society groups. The legislative text includes useful elements—prohibitions on manipulative interfaces and a requirement that companies exercise reasonable care. Yet it avoids establishing a clear horizontal obligation demanding firms assess, document and prevent foreseeable manipulation across their systems.

While the text would help courts and authorities gather evidence on how online commercial practices operate and harm consumers, it still places the evidentiary load on regulators, civil society and affected people rather than companies holding the data. Firms know which designs they tested, what they optimized for and how users responded. They know which moves lifted spending, cut cancellations or prolonged engagement. Regulators, civil society and impacted communities typically must reconstruct that information after harm occurs.

The Commission examined stronger approaches to shifting or reversing the burden of proof. The European Parliament and Council, as co-negotiators, should revisit them. A proportionate framework could require traders to assess foreseeable risks before deploying potentially harmful practices, with documentation of design rationales. For practices already known to pose serious risks, the law could presume them unfair unless the company proves otherwise—without imposing identical duties on a small online retailer and a firm running sophisticated behavioral optimization. If companies can measure exactly how their systems sway us, asking them to demonstrate fairness hardly seems excessive.

Children deserve protection. So does everyone else

The DFA's interaction with the proposed EU KIDS Act (already posing risks to internet safety) adds complexity. Both establish age-dependent protections, including stricter limits on personalized ads and gaming for minors, plus separate safeguards against addictive design.

The DFA does welcome things by restricting extra personal data collection to identify minors. Yet the combined approach may spur further age sorting and potentially invasive, harmful verification. The underlying question persists: why should protection from manipulative practices hinge so heavily on someone's traits, including age?

Children face particular risks, but harmful commercial practices do not become benign at age 18. Comparisons to alcohol, tobacco or gambling sometimes surface in this debate—sectors where consumer choice alone has proven insufficient against harm. Digital spaces, though, are also where people communicate, learn, organize and exercise fundamental rights. Approaches centered on restricting access would carry serious consequences.

A better lesson: foreseeable harms can justify regulating company conduct rather than focusing on who should be barred from services. Drawing lines around harmful commercial practices works better. That means platforms delivering less degraded experiences—benefiting people and their rights while leveling the competitive field. It addresses root causes without making intrusive age verification the price of basic protection.

The Commission's recent preliminary findings on TikTok under the Digital Services Act underscore this point. Authorities are already examining whether engagement-driven recommender systems pose unacceptable risks. The DFA could amplify that work through consistent rules across the digital economy, beyond the largest platforms. Tackling the commercial incentives and optimization driving exploitation beats treating each harmful feature as isolated.

A chance to change who is responsible

The picture is plain: Brussels acknowledges addictive design yet largely relies on opt-outs; it studies situational vulnerability yet narrows legal language; it examines fairness-by-design and stronger evidence rules without clearly translating them into preventive duties. What comes next?

Civil society groups aim for optimism: for once, the Commission has tabled legislation meant to strengthen digital rights, and given the current appetite for deregulation, that warrants hope—however brief. The Parliament and Member States now hold the chance to toughen the proposal and give it real force. Safer defaults, clearer bans, meaningful remedies and proportionate corporate accountability are urgently needed and would make a real difference.

For years, people have been told to read privacy notices, manage consent, tweak settings and defend themselves against systems designed by companies with far greater information and resources. Firms design these systems: they test effects, grasp commercial aims and profit from operation. How much longer should people battle the design of the services they use daily?