privacy

The EU's Digital Fairness Act Is Coming for Privacy Dark Patterns

The EU's Digital Fairness Act targets manipulative design in apps and cloud services. Here's what it prohibits and why it matters for how your data gets collected.

In the third quarter of 2026, the European Commission is expected to formally propose the Digital Fairness Act — a piece of legislation that would extend EU consumer protection into the design layer of digital products.

This isn’t another cookie-consent law. The Digital Fairness Act (DFA) targets the design patterns that sit underneath the consent flows: the choices about defaults, presentation, and friction that determine whether users meaningfully understand and control what they’re agreeing to. It’s regulation aimed at how products are built, not just what policies they publish.

For people who use cloud storage, photo apps, and personal file tools, the implications are significant — because many of the design patterns the DFA would ban are common in exactly these categories.


What the Digital Fairness Act Would Target

A 2022 European Commission study found that 97 percent of the most popular websites and apps used by EU consumers deployed at least one dark pattern — a design choice intended to steer behavior in a direction that benefits the company at the expense of the user.

The Digital Fairness Act would codify prohibitions on several specific categories of dark patterns, with a particular focus on:

Manipulative defaults. Settings that pre-select the most data-sharing option, requiring deliberate action to choose a less invasive alternative. The classic example: an app that defaults to sharing your data with advertising partners, with a small “opt out” link buried in settings.

Misleading framing. Presenting a data consent request in a way that makes “agree” visually prominent and “decline” harder to find, smaller, or grammatically negative (“Don’t use my data”). Design that communicates the “right” answer rather than presenting a genuine choice.

Confusing or staged consent. Multi-step consent flows that start with agreeable requests (“Can we improve the app using your data?”) and only later reveal what “data” means or who receives it.

Urgency and social pressure. Countdowns, “X other users agreed” messaging, or warnings about limited functionality that don’t accurately reflect the technical consequences of declining.

Grief patterns in cancellation and deletion. Flows that are deliberately difficult to complete, require multiple steps that can each be abandoned, or present emotionally manipulative messaging (“Are you sure you want to delete your memories?”) on the path to deletion.

The DFA would also address addictive design — features explicitly engineered to maximize time-on-app at the expense of user autonomy — and unfair personalization, where AI-driven recommendations are tuned to promote company interests rather than user interests.


Dark Patterns in Cloud Storage Specifically

If you’ve used a major cloud storage or photo management service, you’ve probably encountered these patterns without necessarily naming them.

Storage compression warnings designed to drive upgrades. Many services notify users when they’re “almost full” at thresholds far below their actual limit — sometimes sending warnings at 75 percent capacity with messaging that implies urgency. The nudge is toward purchasing more storage, not toward managing what’s already there.

Ambiguous data processing toggles. Settings described with language like “help us improve your experience” or “enable personalized features” without clear disclosure of what data processing that authorizes or who receives the data. Agreeing sounds neutral or beneficial. The actual implications may include training AI models, enabling ad targeting, or sharing data with third-party partners.

One-click consent with multi-step opt-out. Agreeing to expanded data collection requires a single tap in a modal that appears during setup. Undoing that agreement requires navigating to a settings menu, finding the relevant toggle, and in some cases submitting a separate form or waiting for a processing period.

Cancellation paths that use friction. Canceling a subscription or deleting an account often involves multiple confirmation steps, cooling-off messages, and offers of discounts or reduced tiers — each step designed to create another opportunity to abandon the cancellation. Some services require contacting customer support via phone or chat to complete a deletion that could technically be instant.

Encryption claims that exceed reality. Descriptions of data security using language like “your data is encrypted” without specifying whether that encryption protects against access by the service provider. Server-side encryption protects against some threats (external attackers who breach storage infrastructure) but not others (the provider’s own access, government subpoenas, internal access by employees). Marketing that collapses this distinction into “your data is secure” is misleading about the nature of the protection.

These patterns aren’t illegal under current EU law in most cases, though GDPR consent requirements constrain the worst of them. The Digital Fairness Act would create explicit prohibitions with enforcement consequences.


How the DFA Builds on GDPR

GDPR already establishes that consent must be freely given, specific, informed, and unambiguous — and that it must be as easy to withdraw consent as to give it. Dark patterns in consent flows that violate these requirements are already illegal under GDPR.

The DFA is intended to reach beyond consent into design more broadly. It would:

  • Create explicit rules for interface design that don’t require proving a GDPR violation (which can be complex)
  • Apply to design choices that aren’t technically consent flows — like the presentation of cancellation paths or the defaults for non-consent settings
  • Establish specific remedies for dark pattern violations distinct from GDPR’s data protection enforcement
  • Address commercial practices around addictive design and unfair personalization that don’t fit neatly into GDPR’s framework

The two laws are complementary. GDPR governs what data can be processed and under what legal basis. The DFA would govern how the choice to process data is presented to users.


What “Fairness” Means in the Act’s Framework

The DFA borrows a concept from behavioral economics: people make choices based on how options are presented, not just what options are available. Two interfaces that offer identical choices — “allow data collection” or “decline data collection” — will produce different outcomes depending on which option is presented first, which is visually emphasized, what language is used, and what the default state is.

“Fairness” in the Act’s framework means that a company cannot design an interface that exploits predictable human behavioral tendencies to systematically steer users toward choices the company prefers. The information presented must be accurate, the defaults must be neutral or protective, and the options must be equivalently accessible.

This is a more demanding standard than disclosure. A privacy policy that technically describes everything a service does but is presented in 12,000 words of legal text designed not to be read complies with disclosure requirements. It doesn’t comply with fairness requirements.


Who Would Be Covered

The DFA, as currently understood, would apply to services offered to EU consumers — including services operated from outside the EU, as long as they have EU users. This is the same jurisdictional approach used by GDPR, the Digital Services Act, and the Digital Markets Act.

In practice, this means any cloud storage service, photo app, or personal file tool with EU users would need to review its design against the DFA’s requirements, regardless of where the company is based.

For services that want to operate globally with a single interface design, compliance with the DFA would effectively mean applying its requirements to all users — the same dynamic that caused many companies to apply GDPR standards globally when it went into effect.


The Timeline

The European Commission is expected to formally propose the DFA in Q3 2026 — which means the draft text is likely to be published very soon if it hasn’t been already at the time of reading.

After a Commission proposal, EU legislation goes through the European Parliament and Council for amendment and approval. For major digital legislation, this process typically takes one to two years. Once approved, there’s usually an additional transition period before enforcement begins.

GDPR took effect in May 2018 — two years after adoption. The Digital Services Act had a transition period as well. Companies with EU users should expect that DFA obligations won’t kick in before 2028 at the earliest, but the direction of travel is clear and the compliance timeline is not as distant as it might appear.


What Changes for Users if This Passes

If the DFA passes in a form close to what the Commission has proposed:

Default settings become genuinely protective. Services would have to start from a neutral or privacy-protective default rather than building toward data sharing. The burden of consent would shift to the service, not the user.

Cancellation and deletion become straightforward. The multi-step grief pattern for canceling subscriptions or deleting accounts would be prohibited. Deletion requests would need to be processed without friction deliberately designed to prevent completion.

Encryption claims would need to be accurate. Marketing that implies protection the product doesn’t actually provide would fall under the DFA’s prohibitions on unfair commercial practices and misleading framing.

AI personalization would require transparency. Systems that recommend content or configuration options based on AI modeling would need to disclose when the recommendation reflects business interests rather than user interests.

None of this would resolve every privacy concern with cloud services. Server-side encryption would still mean the provider can access your data. Data sharing with third parties for advertising would still be possible with genuine consent. Government access under legal process would still be available to authorities. But the layer of manufactured confusion — the design choices that prevent users from making real choices — would face explicit legal prohibition.


The Signal the DFA Sends

Most meaningful privacy protection doesn’t come from users reading policies. It comes from services being designed in ways that don’t require you to fight the interface to protect yourself.

The Digital Fairness Act represents a regulatory bet on design accountability — the idea that the choices companies make about defaults, framing, and friction are consequential enough to regulate directly, not just disclose and leave to users to navigate.

Services built from the start around genuine user control — ones where deletion is permanent and straightforward, data collection is limited to what the service actually needs, and defaults favor privacy over data sharing — don’t need to redesign anything when the DFA goes into effect. They’re already built for what it requires. The DFA’s burden falls on services that were built the other way: ones that treat user inattention as a feature and friction as a retention mechanism.

That gap between how some services are built and how the DFA would require them to operate is the measure of how much design has been working against users’ privacy interests.

Your memories deserve better than an ad platform.

Try daftei free →
← All posts