When Are Third-Party Browser Engines Coming to iOS? & Fixing the Mobile Web App Ecosystem

• Apple's Browser Engine Ban Persists, Even Under the DMA
• Open Web Advocacy 2025 in Review
• 28% Faster: The Blink Prototype That Shows Why Apple's iOS Browser Engine Ban Must End

Yeah, thanks Bruce and I don't think I'm going to be able to top that and unfortunately my speech is going to be nowhere near as comedic. So I think an important place to start is by asking the question is what sort of web do we want and lots of people are going to have different answers. The web can be many different things to different people and I think one of the extraordinary things about the web is that most people in this room would believe that it's one of the wonders of the modern world. But despite any difference of opinion I'd argue there's a few fundamental properties that we could all agree that we'd want the web to have and that is it should be fast, it should be snappy and responsive on whatever device you have, it should be secure and it should be secure by default without users having to think deeply about it and it should be capable.

The web should provide the building blocks that they need, that developers need to create new kinds of applications rather than deciding in advance what the web is and what the web should be allowed to do and most importantly it should be interoperable and it should work across all operating systems and devices with as little platform specific work as possible and finally it should be open where no single technology company can dictate what people can build using the web and what users can access or demand a cut of somebody else's work simply because they control the operating system underneath it. Now on desktop these goals have largely already been achieved and through a number of different studies one by Microsoft and another one by Intel we think that approximately 70% of all user time is spent in the browser or is powered by the web. So even a bunch of native looking apps like Teams or VS code or even the chatgpt app when you dig into the internals are web based underneath. However on mobile according to the UK's competition and markets authority only 8% of user time is spent on the web. 70% on desktop, 8% on mobile.

So why is the web so phenomenally successful on desktop but comparatively failing on mobile? Can it be the touch screen having a smaller display or being battery powered or having different form factor really explains such a radical gap? Why are native apps so dominant on mobile while the web and web apps are just pushed to the side? And a significant part of the issue is missing functionality, discoverability, performance and stability on the mobile web particularly on iOS. And this has got nothing to do about the capabilities of the devices. The devices we have today are many times more powerful than the desktop computers we had 15 years ago. Now one of the problems we have now is now it fails some of our requirements because now the web is no longer fully interoperable or fully capable. The functionality that I can access for a desktop computer but I can't access it on mobile. And the question is why is this? Because there's lots of capable browser vendors, surely at least one of them can push the web forward on iOS. And so almost everybody here probably likely knows, first 17 years ago Apple made the astonishing decision to ban all the other browser vendors from being able to ship their own browser and compete on iOS. And they did this by rule 2.5.6 which says the rather innocuous apps that browse the web must use the appropriate WebKit framework and WebKit JavaScript. And what that actually meant in practice was that on iOS and iPadOS browsers from Samsung, Oprah, Brave, Vivaldi, Microsoft can't use their own engines that they do everywhere else and these engines take thousands of hours, thousands of engineer hours to develop and they're excluded from Apple's most successful consumer operating system. Competing browsers then can only build shells around a very specific unaltered version of Safari's web view whose component, whose features is exclusively dictated by Apple and that matters for the web and web apps because it means browsers literally can't improve the web on top of what Apple provides and that sets a ceiling on iOS and sets a common denominator on the entirety of the mobile web and this move came alongside its decision to make its app store the only way to distribute native software on iOS allowing it to take a third party cart of all developer work by not competing on merit but by blocking all alternative native app distribution and preventing web apps from becoming the alternative.

So friction kills, a web app you never discover is an app you never install. The main issue is that Apple has gradually hidden away the method of installing web apps. In January 2008, and people forget this, but the first apps that were available on iPhone were web apps. They weren't native apps, that came about a year, a year later. And in January 2008 Apple introduced add to home screen for Safari and it was very prominent. It was literally the middle button, add to home screen. So very, very easy to use interface. And it even created, they even had an entire app library, it was a curated directory of web apps at Apple.com web apps which showcased apps like Facebook, Bejeweled and Sudoku.

And it was at this time that I watched Tim Cook's, not Tim Cook, Steve Jobs' speech where he talked about a great way to build new apps for a mobile device. One that has never been available before. You don't need an SDK, you stick it on your server, you don't have a complicated distribution system. You just stick it on your site and if you know how to write apps with the best web standards then you can build apps for iPhone today. And this is when I first started building web apps. The Apple then launched its App Store for iOS on July 10th 2008 with 500 apps and personally I thought it was a good thing because I thought, well, web can't do everything.

There's always going to be a lag between functionality that comes out for native operating systems and the web's going to take a while to catch up and standardize so having another option for developers is positive. In iOS 6 Apple introduced smart app banners and there's a few examples up there which was basically an integrated mechanism in Safari that allowed websites to advertise their native iOS apps directly from the top of the web page. And by contrast web apps started to be sidelined. We didn't get the equivalent banner or install prompt for them, just the time screen that we had before, which was one of Prev's two options but now was one of nine options. But still, you know, possible to find.

But it was significantly harder to find and install web apps. And then just before WWDC 2013 Apple scrubbed the entire web app library from their website ending their official promotion of web-based alternatives to native apps. By iOS 9 Apple introduced universal links, basically the ability to register a URL from a native app and then make it easier for websites to create direct links into a native app and even prompting users to do so. But by iOS 12 Apple had made the installing web apps even less discoverable. The add to home screen was no longer visible on the main share menu, it was instead buried behind a horizontal list in which users had to scroll to. In iOS 13 Apple updated the share interface once again and it was added onto the screen page of results requiring users to scroll down to find it.

Now they could have stuck it on the first page, they could have done, but they chose not to. And being off the first page of results is incredibly significant in terms of friction and discoverability. Unfortunately we don't have detailed statistics for this, for the impact of moving it from the first page to the second page and each successive hiding of the install mechanism. However positional bias has been very heavily studied and the effects were really large. A really good example was on Google search which found users click on the first position 10 times more than the 10th position and only 0.44% of the searches go to the second page of the search results.

With iOS 14 in 2020 they introduced app clips which were essentially, you know, byte size proportions or portions of an app that could be launched on demand without having to fully install or download it and they were also surfaced via Safari and other methods from other apps and they let the user open the, you know, like a mini version of the app to use the subset of their features immediately and then in iOS 15 they made the add to home screen even less discoverable by even pushing off the first share sheet on the iPad. Then in iOS 16 Apple did a full redesign of Smart Banners and the new install flow didn't even take them to the app store they could just install it directly from the website without having to jump at all to the app store and the entire look of the Smart Banners were redesigned and Apple removed the apps app store that they used to have the stars at the top but they removed that presumably because you stick the stars on there people less like to download the app so that got removed and it got changed to the second version you see there which is pretty much the current one that we have today.

Then in iOS 16 Apple added add to quick note above add to home screen in the share menu pushing it further down the list and in February 2024 they launched their AR/VR headset vision pro and strikingly they chose to remove the ability to install web apps entirely. So really what we have is we have two stories here we have the Safari team making it progressively easier and easier to install apps from Apple's app store in Safari carefully making it more prominent and reducing the number of clicks and interactions the user needs to make while at the same time the same team has been adding extra steps into the web app install flow and making it progressively more and more hidden. What makes it worse is the outright refusal to implement something like install prompts while having essentially implemented two different equivalents for native apps and this matters enormously because the critical part of web apps being a viable substitute for native apps is discoverability.

While all of this was happening, Apple was making the opposite argument to regulators and courts all around the world and this is basically in the countless jurisdictions. They were basically saying don't regulate the app store because developers can always use the web instead and we've got an enormous number of quotes from Apple and even today they're repeating it. For example, they told the ACCC, "Even if a user only owns iOS-based devices, distribution is far from limited to the Apple App Store because developers have multiple alternative channels to reach that user. The whole web is available to them and iOS devices have unrestricted and uncontrolled access to it."

They also said, "Web browsers are not only used as a distribution portal but also platforms themselves hosting progressive web applications that eliminate the need to download a developer's app through the App Store or other means at all." For the App Store rules, the opening paragraph states, "For everything else, there is always the open Internet. If the App Store model and guidelines are not the best for your app or business idea, that's okay. We provide Safari for a great web experience too." So Apple spent years holding up the web as the competitive alternative to the App Store while simultaneously denying it the features and discoverability it needs to actually compete.

And by banning competing browser engines on iOS, they've prevented anyone else stepping in and building the alternatives themselves. So how do we fix this? Fixing it is actually really simple. We just need other vendors to be able to compete globally on iOS with their own engines. We need Safari to implement a discovery mechanism for web apps at least as good as the one they have done for native apps. And this is really important because the success on the web on mobile should not be contingent on Safari's demise. In fact, we want the opposite.

We want the web to be successful and we want Safari to be successful because they're one of the major competitors in the browser space. We need third-party browsers using their own engines to be able to manage and install web apps. And we need browser vendors to have access to the operating system features that they need to compete and that they need to let web apps compete with the native app ecosystem. And browsers should be technically able and allowed to provide web app feature parity with native app and the web shouldn't be banned from having access to functionality it needs or being provided with or only being provided with inferior versions. And the reason I say this is simple is we basically have this on every single other operating system iOS is the only holdout.

So we started OWA really because of a lack of progress on iOS and somewhat really naively our first instinct was let's just ask Apple to fix it and we simply got ignored. They forced us to ask a different question which was why isn't Apple fixing this? And the more we looked at it the incentives became very very clear. So first up there was a 20 billion dollar a year search deal with Google which as crazy as it sounds is almost all profit making Safari almost certainly Apple's highest margin product out of anything. In 2020 Google's payment alone was equivalent to astonishing 17.5% of Apple's entire operating profit and fair browser competition on iOS would directly threaten the value of that deal.

Second there was the Apple's App Store which generates roughly 24 billion a year for Apple and competitive web apps would at minimum cut into that revenue stream and threaten the App Store's position as the mandatory distribution layer for software on iOS. We realised that if browser competition, web app competition were ever going to be opened up properly on iOS it was with competition regulations and that's why we founded OWA. Now we've been working for five years now, we've been talking to competition regulators all over the world including the UK, the EU, the USA, Japan and Australia and a lot has changed and we have had a fair bit of success. The first part Safari implemented an initial version of push notifications. It's missing a whole bunch of features compared to the native ones but it's a really good start and we're also confident that because of our work Safari's budget got significantly increased which is really good to see.

And just so we're clear to everyone in the room, we're not anti-Safari. If Apple bumped up Safari's budget to a billion dollars a year, a mere 5% of their revenue that they collect from Google, that we would be overjoyed and that would be fantastic because that would really start to push the web forward. In the EU they passed the Digital Markets Act which explicitly bans gatekeepers like Apple from imposing browser engines with the purpose of preventing them from holding back the functionality of third-party browsers and the web applications that they power. In Japan they passed the Smartphone Act which imposes a similar requirement but only for iPhones. And in the US the DOJ opened up an entire antitrust case into Apple and one of the core issues raised is Apple's ban of third-party browser engines and how this prevents web apps from being a viable competitor to their app store.

And finally in the UK we had a market study, we had a market investigation referenced into browsers which found a huge list of claims against Apple and then we finally got the DMCCA which is the UK's equivalent of the EU's Digital Markets Act and the CMA is currently investigating Apple's restrictions on browser engines and the potential of web apps. Now the first one that came into force was the DMA and that was in the 7th of March 2024. So on that date Apple had to allow third party browsers to fairly compete on iOS with their own engines. That was two years ago. So the question is where are the browsers? Unfortunately Apple has put up a number of barriers to prevent it being successful in practice.

Apple could have when this came out said well we essentially banned the competition for the last 17 years, it was a good run, alright let's let everyone compete now. They could have simply allowed browser competition and allowed it globally. That was a genuine option. But what they chose to do was construct a series of barriers and artificial complications. One that allowed browser competition to be technically allowed but completely unfeasible for any browser vendor in practice. Now there are a huge list of issues and I'd need to run this for at least about six hours to go through them all with you. But just to give you a sense of the big ones.

The first thing they demanded was that browser vendors create an entirely separate app from their existing app and try and cannibalize their own users and migrate them over. They then outright refused to let browser vendors install web apps powered by their own browser engine and as many people here in the room probably remember, try to remove web apps altogether to stop this happening in early 2024. Apple prevents browsers from using their own engine when EU users travel basically at a switch. So the second you leave an eligible jurisdiction you're expected to instantly toggle back to the WebKit version. Now for those that might not be aware the data storage in one browser is completely different from the data storage in another and porting all that data at a drop of a hat is bordering on next to impossible.

And then the next one is Apple won't let web developers test their web apps and websites unless they're physically in the EU or actually have to fly in to do the testing. And even more ridiculous, it's fixed now but for the first six months they wouldn't even let browser vendors test their own browsers outside the EU. The next one Apple has imposed an API abstraction layer called browser engine kit on all the browser vendors which is missing, it's been very buggy, it's missing lots of features and it's had a great deal of stability issues. They could have just said use the existing APIs that are available on iOS and browser vendors could have built their browsers but they were locked to this abstraction layer. And then Apple's locked by technical design the system provided in app browser to Safari rather than users default browser.

Then they created an extremely aggressive set of legal contracts to be able to get the required APIs which grants Apple the right to remove or break any API, forces the vendor to follow every rule in the unbounded and unspecified number of documents basically on star.apple.com and it allows Apple to in the case of what it considers a breach in its sole opinion remove not only the browser vendors EU browser but every app that vendor has from every operating system that Apple controls globally. So now we understand the competition backdrop and the issues back to web apps and I think it's really important to think about what happens if web apps become an effective substitute for native apps. Suppose an iPhone web app could reliably have excellent performance and stability, proper installation minimal friction, good discoverability, robust and complete push notifications, all of the device APIs, sophisticated media APIs, deep OS integration, universal links and suppose Google, Vivaldi, Microsoft, Mozilla and others were free to implement any web capabilities in their own engines without having to wait for WebKit.

Then a large category of applications suddenly become open URL, install web app, use app. Then each time you update, you just update the code on your server rather than join the Apple developer program, comply with app review, package up a native app, submit it, distribute it for Apple's system, and then follow all the app store rules and share 30% of your revenue with Apple. Then every time you go for an update, you have to go for this complex, slow and frustrating app store review process. And as one regulator asked us, "If web apps could all do that, why do we even need app stores?" Now it wouldn't be a 2026 talk without discussing AI, and we think it probably makes the issue more important, not less.

So obviously it makes the development of web apps and native apps faster and more accessible, particularly for those with no or lower coding skills. And this isn't fantasy, I've personally got AI to make web apps that would have taken a small teen months, and these are mainly apps for my children to use. And even just for this trip, I made an entire meeting management app, bespoke, just to help manage all the information for this trip. What this will go through is it will suddenly bring a sharp contrast to all the disadvantages of having to go through an app store instead of using the web. Imagine we had real browser competition everywhere and really good support for web apps everywhere. You tell an AI, make me a scheduling and membership application for my local sporting club.

A few minutes later it generates a reasonably polished application and then you open a URL on your iPhone and the browser says do you want to install this app, you tap it and you get an icon, you get notifications, you get offline storage, you get background sync, file system access, camera access, full GPU acceleration and whatever modern web standards provide. There's no Xcode project, there's no Swift, there's no app store submission, there's no Apple review, there's no $99 membership, there's no platform specific binary and there's no need to build one for Android and for iOS. And the developer can update it simply by changing their server. And then for small bespoke software, so you think clubs, businesses, internal tools, events, prototypes, personal software, they're unusually well suited for this model.

And the combination of mature web app support and frictionless installation could fundamentally change the economics of the market. And native app stores have benefited enormously from the fact that building applications is expensive. If an application creation becomes almost free, then having to package, sign, review and distribute every tiny piece of software for a centralized app store becomes the real burden. Not to mention having to share your revenue with them. So what now? Well, the good news is the issue hasn't been forgotten.

Regulators are very aware of the barriers that are put in place. We have three opportunities to fix this. First, the UK can impose conduct requirements under the DMCCA. And essentially what this is, is a set of bespoke rules for Apple to allow browser vendors to use their own engines, have the APIs access they need, and to be able to install and manage web apps with their own engines. And both the CMA's previous investigations, the market investigation reference, and the SMS investigation proposed doing exactly that. Even in the legal explanatory notes for the DMCCA, it says it directly uses browser competition as the example, as the sort of issue the CMA could fix with it.

Secondly, under the Digital Markets Act, the EU could open up a specification proceeding into Apple. Basically, this would be the Commission telling Apple in very precise terms exactly how to comply with the DMA. And some of you might have heard of the recent third-party device API specification proceeding, which among other things gave us interoperable airdrop between Android and iPhone. But the EU handed down a big list of major requirement changes to iOS and Apple did every single one of them to the letter. And the final one is the DOJ case in the USA. And this issue is directly listed in the case and the DOJ wins, thanks to it.

And they have, and the DOJ have the power to implement global change and trust is very slow but it's super powerful and the latest news we have is Apple is looking to settle. And the important thing is we only need to get this working in a single jurisdiction, in a single market, and then Apple's barriers will collapse globally. And if we fix this, it will be amazing. We get real competition with app stores, we get cheaper and higher quality software, we get a truly interoperable platform where developers can build once and reach everyone, one that is free of the gatekeeper tax, one that is more secure, more private and more capable. We can restore the web as a platform that works for everyone, everywhere, on every operating system.

So before I finish, I'd just like to say thank you to everyone that makes our work possible, everyone that volunteers, everyone that pushes us for change, all of the various browser vendors who are helping us fight this fight to try and push the web forward, and to the regulators who have taken our arguments very seriously on a topic that is quite technical, esoteric and abstract, even if the harms to consumers and businesses are very, very real and they're definitely trying to solve this issue. So the final message I'd like to leave you with, don't give up, we're coming to the finish line, we just need to push it over. Thank you.