BONN, 20 AUG 2026 — Germany's competition regulator has closed a four-year investigation into Apple's App Tracking Transparency framework after accepting binding commitments. Apple will redesign the consent prompts so that requests for its own advertising look comparable to the ones third-party apps must show.
The complaint was never that Apple asks for consent. It was that Apple asks differently for itself.
The settlement
The Bundeskartellamt found that Apple made it easier for users to consent to its own advertising than to that of competitors. Since iOS 14.5 in April 2021, third-party apps have had to ask permission before tracking users across other apps and websites for personalised advertising, while Apple's own services were not subject to the same restrictions.
Under the commitments, Apple will remove possibly discouraging symbols and wording from the prompts, align the design of requests for its own services much more closely with those shown for third-party apps, and allow publishers to combine the Apple-mandated prompt with other consent requests they must show under data protection law.
Four years is worth registering as well. The case opened in June 2022, roughly fourteen months after the framework shipped, and closed this month. Whatever the merits, an advertising market restructured itself completely inside that window: measurement moved to aggregated reporting, budgets shifted toward platforms with logged-in users, and a generation of attribution companies went out of business. A remedy arriving now equalises a prompt in a market that already reorganised around the old one.
The asymmetry, stated plainly
App Tracking Transparency was defended as a privacy feature, and it is one. Users got a clear choice about cross-app tracking. Most declined, and a large amount of behavioural advertising infrastructure stopped working.
The competition question is narrower and does not dispute any of it. A company that both sets the rules for a platform and competes on it can write rules that apply to everyone and land differently. Third-party advertising required a prompt with a design Apple controlled; Apple's own advertising operated under a different arrangement with a different interface. While both arrangements may be lawful in isolation, a competition authority examines how they work in combination.
The remedy reflects that precisely. Nothing here weakens the consent requirement or restores third-party tracking. It equalises how the question is asked, which is the only thing the regulator claimed was wrong.
The wording change is the substantive part
Removing discouraging symbols and language from a consent prompt is not a cosmetic change.
Consent rates are extraordinarily sensitive to presentation. The order and colour of buttons, the phrasing of the negative option, or the presence of a threatening-looking icon can all move outcomes by amounts that would be considered enormous in any other field. An interface designer can produce almost any consent rate they want from the same underlying question, and everyone in the industry knows it.
That is why the finding is about design rather than policy. Two prompts asking materially the same question can produce opposite results, and if one of them is yours and the other is your competitor's, you have set the market share without changing a rule.
The wider implication reaches every consent interface, including the cookie banners readers see daily. If a regulator will treat prompt design as conduct rather than presentation, a great many banners built to discourage refusal are exposed to the same reasoning.
Apple's position, while not accepted by the regulator, is coherent. A platform owner arguably does know things about its own users that no third party should, and holding first-party data to the same disclosure standard as cross-app tracking treats different activities as if they were the same. The regulator has not accepted that argument here, and it is not an unreasonable one. What made it hard to sustain is that the difference showed up as interface design rather than as a stated policy anybody could evaluate.
Combining the prompts is the change publishers will notice
The least-covered commitment has the most immediate practical effect. Publishers may now combine the Apple-mandated prompt with the separate consent requests required by data protection law.
Until now a European app had to show two consent dialogues in sequence — one required by Apple, one required by law — asking overlapping questions in different formats with different wording. Every additional dialogue costs completion, and users faced with a second request having just answered a first frequently dismiss both.
Merging them removes a pointless friction. It is the kind of fix that only a regulator can produce, since no single company had authority over both prompts.
Why this reaches beyond Germany
Apple will not maintain a separate consent design for one national market. Interface changes of this kind ship broadly, which means a German competition decision will very likely alter what users see in Singapore, Manila and Jakarta without any authority in those places acting.
We made the same observation yesterday about the 29-state trial against Meta: platforms ship one product globally, so a remedy ordered in one large market becomes the default everywhere. It is convenient for users in markets with less regulatory capacity, and it means the effective standard for consent design across Southeast Asia is being set in Bonn and Oakland.
There is a second-order point for regional publishers and app developers. The economics of mobile advertising here changed sharply when App Tracking Transparency arrived, and any material change in consent rates changes them again. A developer monetising through advertising should expect the numbers to move and should not assume the direction.
What we could not establish
What the redesigned prompts will actually look like, and when they ship. Commitments describe outcomes rather than interfaces, and the difference between compliance and substantive change is entirely in the implementation.
It also remains unestablished whether the changes apply only in Germany, across the EU, or globally. Other open questions include the deployment timeline, how or if the regulator will measure consent rates, how the combined prompt will work, and whether other competition authorities will follow suit.
What to watch
Watch the consent rates. If third-party opt-in rises materially after the redesign, that retrospectively measures how much the previous design was doing, and it is the number the whole case was really about.
Then watch whether other regulators cite this. A binding-commitments settlement is not a precedent in law, and it is a very usable template, particularly for authorities examining the same conduct under digital markets rules.
Finally, watch what happens to cookie banners. The reasoning applied here is that discouraging design is conduct, not presentation. That argument has been missing from consent-banner enforcement, and it is now sitting in a decision anyone can read.
View author profile → · Editorial policy
