Context: The European Union’s Digital Markets Act (DMA) will not open up the market iOS apps anytime soon as Apple has promulgated a new set of EU-specific rules that, in conjunction with various technical measures and pre-existing app guidelines, render the DMA largely ineffectual. Enforcement of the current DMA must be tried, but it will likely be slow, Apple will leverage its resource advantage, and the outcome will turn on the level to which the EU judiciary is prepared to take the concept of purposive interpretation (February 19, 2024 games fray article).
What’s new: In this article, games fray outlines a set of changes that EU lawmakers could make to the DMA. Some of those measures may be generally desirable to make the DMA achieve its objectives, but at minimum they are needed to up the ante for Apple (and, if necessary, Google) in the app distribution context.
Direct impact: It is obvious that the EU Parliament is soon going to be in reelection campaign mode, but the European Commission as well as the EU Council’s experts could start their work anytime. The proposed modifications and clarifications would hugely increase the likelihood of Apple’s current and future countermeasures being held unlawful, thus potentially resulting in huge fines and substantial damages. They would also accelerate the process, giving EU-based iPhone users and the global app developer community serving EU customers (“business users” in DMA lingo) the benefits of open app markets at an earlier stage. And they would simplify the task of the DMA enforcers at the EC’s Directorate-General for Competition.
Wider ramifications:
- What’s at stake is the ability of the EU to deliver results that are actually felt in the marketplace, as opposed to talking the talk without walking the walk. The EU will make itself ridiculous if the only tangible benefit of the DMA is that EU-based users get native iPhone “adult content” apps.
- These proposals may also be a useful contribution to the debates in other jurisdictions working on digital platform laws, particularly the UK and Japan.
- For Apple, the EU represents only 7% of global App Store revenues, but what Apple primarily fears is that if the EU proves open app markets aren’t inherently insecure and don’t compromise end users’ privacy, Apple will run out of arguments to dissuade policy makers in other world regions from taking action.
The proposals below have a certain logical order, but that does not imply any prioritization. In the end, what is needed is a coherent set of rules. This article doesn’t propose specific wordings, but explains the gist of the required amendments. Furthermore, this article doesn’t discuss whether some of those rules should also apply to other parts of the DMA than app stores, but that is just in the interest of time.
1. Confederated App Stores: clarification regarding platform designations
Despite all of its criticism of Apple’s walled-garden policies, games fray does not dispute Apple’s access to justice and its right to test, within reason, the limits of the DMA and similar regulations (or court rulings). In games fray‘s opinion, Apple is being confrontational and partly destructive or evil, but not engaging in lawless behavior.
Apple does have a point that the EC’s designation of all Apple App Stores (even including macOS) as a single core platform service is debatable. Apple is, therefore, reasonably appealing, but its stance on the App Store for the iPad is nonsensical. At least with respect to the iPad, there can be no doubt that it is the same App Store as for the iPhone. In fact, the author of this article has personally uploaded apps to Apple’s App Store, and while one is free to make special iPad versions of apps, most apps simply use the same executable code on both types of devices. And a consumer who purchases an iPhone app can use it on their iPad (and vice versa).
It is highly unlikely that Apple can avoid that the DMA also applies to the App Store with respect to the iPad. The courts are not likely to be fooled into making a distinction. With the Apple Watch it’s a little bit less clear because one doesn’t run identical apps, though many Watch apps are simply companion apps to iPhone/iPad apps.
That litigation over platform designations will take years. In the meantime, the fact that app makers would have to distinguish between iPhone and iPad apps in the EU (even in respects in which they don’t have to differentiate anywhere else in the world) will constitute yet another disincentive from distributing apps through alternative app stores.
The right response to Apple’s litigation and its obstructive conduct with respect to the iPad is legislative clarification. Therefore, games fray proposes that an amended DMA specifically state that
- if an app store is designated as a core platform service, Confederated App Atores must be treated as one core platform service,
- defining Confederated App Stores based on such criteria (any single one of which would be sufficient) as
- cross-purchase (buy on iPhone, use on iPad) and/or
- largely overlapping application programming interfaces for the underlying operating systems (e.g., more than 25% of the API functions of one operating system are also provided by the other operating system) and/or
- largely overlapping customer bases (for which a threshold could be defined; e.g., if more than 25% of the users of one version of the app store also use another that is part of the same confederation of stores) and/or
- any other indicia (yes, the language should be broad and inclusive) of a close technical and/or commercial relationship.
2. Most Favored Nation Requirement: closing any loophole for disadvantaging EU-based users
Apple’s decision to hobble Progressive Web Apps (PWAs) specifically in the EU (February 13, 2024 games fray article) adds insult to injury. Such reckless and arrogant conduct provokes and warrants a forceful response.
The DMA could come with the legislative equivalent of what many commercial agreements have: a Most Favored Nation (MFN) clause. In Amazon’s case, that has raised competition concerns, but that is a non-issue here.
The EU has the right to tell someone like Apple that if they want to sell in the EU market, they must not deprive EU users and the “business users” serving them (i.e., app makers) of any choice or procompetitive benefit they offer anywhere else in the world. In the current context, that means PWAs. As games fray explained in another article, Apple needs PWAs as a defense to (conventional) antitrust claims in other jurisdictions, but feels that in the EU it can argue that the availability of alternative app stores obviates the need for (properly supporting) PWAs. That is, by the way, similar to Apple’s approach to “sideloading” (a pejorative term for direct installations of apps): Apple says that if you have alternative app stores (and can directly install from the web the front-end apps of those stores), that’s all you need to have one (and only one) alternative to Apple’s App Store (January 26, 2024 games fray article).
The definition would have to be reasonably inclusive. A good test case is the ability to run PWAs in full-screen mode. That used to be possible until Apple removed that feature (even from its own Safari browser!) as part of its DMA response. It’s a technical feature, but it does have competitive implications. As games fray explained, most browser games aren’t really usable except in full-screen mode as they’d look ugly with browser controls on the screen, need all the screen real estate they can get, and unless it’s a game like chess, users will have to interact with the user interface at a high pace, which would inevitably result in them hitting browser controls by accident, thereby disrupting their gameplay.
The EU with its huge market (even if Android is way more popular there) can simply demand that any procompetitive option that exists on iOS anywhere in the world must also be offered in the EU. If other jurisdictions did the same, Apple would soon have to open up on multiple fronts on a global basis. Whatever they do in jurisdiction A because of a class-action settlement, in jurisdiction B because of a digital platform law or in jurisdiction C because of a regulatory order would then also benefit everyone else.
Optionally but desirably, the MFN requirement in the DMA could be made even more inclusive by referring not only to procompetitive choices offered by Apple elsewhere in the world at a given time (geographic vector), but also whatever procompetitive choice they offered in the past (including in the EU) within, say, the last 10 years (temporal vector). Unless this is made a one-way street, it will be a “one step up, two steps back” situation where Apple will comply with one new rules and then take one or more countermeasures such as what is happening now in the PWA context.
If the argument is that some restrictions of choice compared to the past are needed because of technological change, then there has to be a burden of proof, with an exacting standard (“beyond reasonable doubt”), on the gatekeeper.
3. Anti-Fragmentation Rule: one app to serve the whole world and across all app stores
For quite some time, Apple has now been playing the game (first in the Netherlands in a dating app context) that app makers seeking to benefit from new choices (such as alternative payment systems) need an “entitlement” for their app, but the entitlement (which is just a technical data point that flags the availability of an option) is available only per app, meaning that different apps with different IDs are needed in different markets.
That kind of fragmentation makes it extremely unattractive for app developers to avail themselves of new procompetitive options, especially in light of the fact that the EU accounts for only 7% of App Store revenues. App makers are then forced to decide whether to migrate all of their users in one jurisdiction to a new app (same name, but new app ID and therefore no automatic conversion) in order to get the benefit of local procompetitive rules or whether to migrate everyone else. For example, if an app has most of its user base in the U.S. and its maker wants to benefit from Epic’s anti-anti-steering injunction, the decision may to be just update the U.S. app and hope that users in the rest of the world will go to the App Store (or an alternative app store) and reinstall the app from there. There will inevitably be a reduction of the user base, and many users will consider it a hassle, which may adversely impact the app maker’s reputation even though it’s all Apple’s fault.
Similarly, Apple forces app makers to decide whether to distribute an app either via its App Store or via third-party app stores, and create friction in that regard.
An anti-fragmentation rule (which happens to be how Google defends some of its Android app rules in a dispute with the EC) would say that the gatekeeper must allow app makers to offer one app in all parts of the world, and on any number of app stores, without requiring separate app IDs in order to benefit from any local rules.
This, again, is a rule that other jurisdictions than the EU should also consider.
4. Anti-Friction Rule: expanding on the Anti-Fragmentation concept
While it should be obvious, Apple’s behavior suggests that the DMA must make it clear that circumvention of the DMA by the introduction of friction (such as unnecessary steps required by users to activate features or the need to uninstall and reinstall apps) can be deemed a violation and give rise to hefty fines.
5. Technical Parity Requirement: pre-empting performance and user experience degradations
So far the discussion is all about issues caused by Apple’s new rules. The PWA situation has now drawn some attention to technical measures, such as Apple disabling browser choice for PWAs and simultaneously degrading the user experience of web apps even in its own Safari browser (where there can be no security justification, particularly since no app is less secure in full-screen mode than in a browser tab).
Technical parity has recently also played a role in connection with the merger reviews of Microsoft’s acquisition of Activision Blizzard.
Technical parity should be part of the MFN requirement discussed further above, but must also be ensured with respect to, for instance, alternative app stores and browsers.
6. Balanced User Communications: enable app makers to contradict Apple’s “scare screen” messages
One of the oldest European principles is audiatur et altera pars: may the other side be heard as well.
The problem with Apple’s “scare screens” when app are downloaded from alternative stores or alternative payment methods are used is that only Apple gets to take its position that everyone other than Apple is by definition untrustworthy. That is ridiculous. There’s no reason to assume that an Amazon, Microsoft or Epic Games (whose games store is one of the largest in the world) can’t operate an app store in responsible ways, or that payments made via Paypal or Stripe are inherently insecure.
One simple, effective and proportionate measure would be to make it a requirement that whenever Apple displays a scare message, it has to juxtapose it (parity in terms of prominence) to the app maker’s message.
So if Apple essentially says “don’t install this app because it can’t be trusted”, there’d also have to be a message where the app maker can say “We’re [company A], you know us and you know you can trust us.”
Optionally but desirably, it could be clarified in the overall context of privacy and security that the gatekeeper must prove the proportionality of a privacy and security measure beyond reasonable doubt.
7. FRAND: the biggest financial issue, and Apple’s own positions can be held against it here
Sometimes regulators just say FRAND (fair, reasonable and non-discriminatory terms) because they have no other answer, can’t agree on one or don’t dare to take a clear position. The question is then how both sides (the beneficiaries of a FRAND licensing obligation and a party that is under such an obligation) can enforce their rights in practice. Timing can be of the essence, and the DMA is about curing market failures not just in ten years’ time but well before hat time frame.
The DMA doesn’t say that gatekeepers are not entitled to any compensation for the use of their intellectual property. That would be a form of expropriation unless justified (which wouldn’t be easy to do, also in light of international treaties) in a wider context. It just says FRAND.
As the European Parliament’s DMA rapporteur, Andreas Schwab MEP, recalled on X (formerly known as Twitter), his committee would actually have liked to give the European Commission the authority to make a determination of what terms are FRAND:
While games fray agrees with Mr. Schwab (and it should be easy for anyone to agree in light of Apple’s new EU-specific rules) that there should have been more in the original DMA than a general FRAND requirement, an EC-led FRAND determination process would amount to governmental (though appealable) price-setting.
It’s possible that too few of the people working on the DMA (in the EU institutions as well as on the side of stakeholders advocating open app markets) were deeply familiar with how FRAND works in connection with standard-essential patents (SEPs). That context is highly important because it’s one in which Apple takes the position that third-party IP (such as patents on the 5G cellular connectivity standard) is not very valuable and that Apple should be free to use other parties’ IP for many years with impunity until it may ultimately agree to pay what a court considers a FRAND rate, and even that only after actual infringements have been proven and the relevant intellectual property rights have survived validity challenges. That is the very opposite of how things would work under the current DMA, where Apple can just require app makers to accept its commercial terms, which in turn means that alternative app stores (except for special use cases such as “adult content”)
The EC has tons of documents from Apple (back from when Apple complained of what Samsung and Motorola Mobility were doing with their SEPs in the early 2010s, but also from policy debates in the 12 years since) where Apple takes the position that it should, for instance, be able to make and sell 5G phones even if a 5G patent holder may not get paid for 5 or 10 years. Apple has been consistent about that, and vociferous.
It’s also worth nothing that U.S. litigation showed Apple spends only about 1-2% of its sales revenues on SEP licenses. One of games fray‘s sibling sites, ip fray, discusses (among other topics) SEP litigation and the related policy debates. There are many thousands of patent families that have been declared essential to the 5G standard, while Apple in its U.S. dispute with Epic Games merely pointed to a couple hundred relevant patents and unspecified copyrights.
Far from engaging in expropriation, the EU could say in an amended DMA that
- the platform owner (i.e., Apple) is not entitled to compensation from alternative app store operators, app makers to the extent they distribute through alternative app stores, and app makers to the extent they distribute their apps directly (web, no app store) to users for compensation beyond a FRAND royalty for its IP (i.e., in these scenarios, Apple could not charge for its access to customers);
- the platform owner must not block access or pursue injunctive relief (sales bans etc.) if an app maker or app store operator enters into a binding commitment to pay a FRAND compensation for the use of Apple’s IP as the final outcome of a litigation because all appeals have been exhausted, an appeal has been waived or the parties have reached an agreement; and
- in order to obtain such compensation, Apple must (as it demands from cellular SEP holders) first establish in court that it actually holds valid intellectual property rights that are infringed.
The courts could be authorized to order app store operators and app makers to put a certain amount of money into an escrow account to protect Apple against the possibility of a debtor’s bankruptcy. The amount would have to be determined by the courts and should not be prejudicial with a view to the ultimate royalty determination.
In practice, this means Apple would have to show that app store operators and app makers actually use and infringe any valid IP. Most patents don’t survive validity challenges in this field. And many patents don’t have as broad a scope as their owners argue in the infringement context. It’s possible that Apple will have no relevant patent at all, or maybe just a handful or a dozen.
As far as copyrights are concerned, Apple’s problem is that API copyrights tend to be weak now. There’s a strong interoperability privilege in EU copyright law, and in the U.S. Oracle lost the case against Google in the Supreme Court because of a fair-use holding.
That leaves the question of what Apple can charge for its developer tools. Those are protected by IPRs far more robustly than its APIs. Here, the solution is the MFN rule proposed further above. If Apple wanted to raise the costs of its developer tools, it would have to do so worldwide and couldn’t do it in the EU alone. If one wants to even close another potential loophole, one could make it clear that developer tools must not be licensed by gatekeepers on an ongoing royalty basis but for a periodic or one-time lump sum.
Apple would have serious problems with a FRAND determination by the courts of law, not only because it wouldn’t ultimately be able to show a whole lot of IP (if any) that is valid and infringed, but also because its own positions on FRAND are so well-documented.
Plus, one could fine-tune some details
The above seven items are the most important changes to contemplate. In addition, some details could be clarified, such as that direct installation from the web must be allowed for all apps (and not just for the frontend apps of alternative app stores).
The purpose of this article was just to provide food for thought and show how the problem could be solved. It would teach Apple a lesson if its conduct led to a far more forceful and more easily enforceable regulation. Enforcement of the current DMA against Apple is not going to solve the problem with any reasonably high probability. And jurisdictions like the UK and Japan now have the chance to learn from what went wrong in the EU before they find themselves in DMA-like trouble.