Google I/O '25: My personal take on the announcement regarding the Android ecosystem for this year and the next
Well, another Google I/O has come and gone, and the Android track for 2025 certainly didn't skimp on announcements. I've watched the presentations (most of them), scribbled down the highlights (and a few lowlights), and now it's time to distill what this means for us developers and the ecosystem at large, while presenting some compliments and criticism I believe to be valid and appropriate. So, grab a coffee, and let's dive in the news!
Part 1: Android 16 - The middle eastern dessert (Very tasty, actually)
Android 16 is the most antecipated announcement, and this year, it made an early appearance. Google’s team even color-coded the changes by importance, so we may know where to focus first, and where to focus later on (red, yellow, green – a bit like a traffic light for our development priorities).
Kicking things off, Jetpack Compose continues to be the solid library all modern Android UI development is now based on. We're getting new modifiers (Autofill text, Autosize text, Animate bounds, and fancier Visibility trackers), alongside welcome performance boosts and a rather impressive 32% reduction in experimental APIs, with increased stability. The way I see, that's a win. Navigation 3 is also on the horizon, rebuilt with Compose in mind, promising slicker transitions and customization, which is, of course, very welcome to have. But the bigger surprise to me was a Jetpack Compose based Media 3 and CameraX new libraries, a sort of overhaul of the already present libraries, and also very welcomed to have available. It seems good winds are blowing Compose further into the future.
Kotlin Multiplatform (KMP) enthusiasts, fans, skeptics and community alike, rejoice! Full Jetpack support for Tier 1 platforms (Android, iOS, JVM) is now available. This is genuinely good news, and shows KMP also has a promising future in the highly competitive market of hybrid mobile development.
Regarding security and privacy:
- Digital Credentials Verification and restoration via new API calls look promising. Frankly, anything that makes the apps safer is welcomed, no questions asked.
- The Privacy Sandbox now lets you isolate any third-party code in its own runtime process, not just ad SDKs. That's a neat trick for containing rogue code, although checking the reputation of a library before using it is also important.
- Android Advanced Protection Mode is another new user-toggleable feature users can set, and that we'll need to integrate with in our apps before they will be able to use it – It means more coding, debugging and testing for us, but hopefully more peace of mind for users.
- And the one I'm particularly glad to see: Theft Protection with Identity Check! An extra biometric layer to deter those snatch-and-run phone thieves? Excellent! We can't have enough of those, folks.
Health Connect is expanding, aiming to consolidate more health data for better medical context – think vaccines, allergies, alongside the usual sleep and movement. A positive change, I believe.
On the performance front, the perennial battery vampire, wake locks, are still an issue, often due to long-running jobs. Google is tweaking Expedited Work in WorkManager to try and tame this beast. Yet, it remains to be seen how effective that will be when applied.
And then there’s the big push: Adaptive Apps -> Android is doubling down on making apps look good everywhere. For larger screens, Android 16 will apparently start ignoring your carefully set orientation and resizability restrictions on activity recreation. While reorienting and resizing is something that should always be considered on our checklist, I fear developers will resist adopting this library and adding proper adaptative code.
Material 3 Expressive is also part of this, an "expansion pack" (as I see it) for Material 3, bringing 15 new/improved components with unexpected shapes, new physics, and colors. It sounds fun, but let's hope it doesn't carry any serious letdowns.
Other notable mentions for Android 16:
- Live Updates: A new notification template for time-sensitive tasks like food deliveries, or map update for navigation. It is a notification in the notification tray changing content every few seconds. Very useful in some contexts.
- Jetpack Glance 1.2 Alpha: A widget building API based on Jetpack Compose. Again, it's only fair for this to be available, since Compose is being standardized as the recommended way to develop UI on Android.
- Google Low Light Boost: A software library using ML for better snaps in the dark. Nice.
- Preload Manager for Media 3 and Native PCM Audio Offload: More tools for media-heavy apps, allowing performance improvements.
- Gemini Nano Integration: Via a new GenAI API and Gemini Live APIs (text/audio I/O), now Android apps may make use of an on-device generative AI capabilities. As a developer who is very interesting in connecting people and integrating software, I find this feature very promising and very interesting.
My Sentiment on the changes for Android 16: Look, performance and usability optimizations in Jetpack and other libraries are always welcome, no questions about it. The anti-theft layer is something that warms my heart, knowing someone took their time to code something to help you in your moment of vulnerability, just after a theft. Even if you don't recover your device, it will likely still protect you by giving you more time to take measures. The way I see it, it's a massive win.
More KMP and Compose-centric libraries like Glance, Media 3, and CameraX are also desirable, and to be expected. I also agree the way Android manages screen reconfiguration is outdated, unnecessarily complicated, and sometimes, a letdown to developers. However, I find it too convenient for Google to push an update for multiple form factors on Android while at the same time, announcing the release of Android XR. This push for adaptive screens across multiple form factors... It is good on paper, but it seems people forget it will come at a cost for companies. And I can't help but think someone might be trying to strong-arm the Android community into supporting the fledgling Android XR... Adapting your app for yet another device category costs real money, and not every app needs to be on every screen.
Part 2: Polishing the Shelves of Google Play
The Google Play Console also got a fresh coat of paint and some new tools under the hood. The app dashboard has been redesigned to surface key metrics like "Test and release" and "Monitor and improve" more prominently. Each core object now gets an overview page, which should simplify navigation.
To tackle the challenge of multi-platform optimization, the console will now flag incorrect edge-to-edge rendering in a pre-review check. A new panel for quality issues and recommended fixes has also been added, alongside an Android Vitals metric for low memory app closures and excessive battery drain. As long as the UI and UX design can keep up with the amount of data the console provides, those changes should be helpful, I believe.
Store listings get more customizable, and a new panel for engagement metrics (installs, ongoing engagement) has arrived. Monetization sees more currencies and payment methods, with an AI model suggesting the most appropriate one. Good.
On the security front, the Play Integrity API gets enhancements for hardware-based security to reduce spoofing and a new abuse detection feature for apps in trial mode that doesn't rely on device IDs, which is a privacy-conscious move. Again, I may be biased, but anything related to security is a win in my book, so kudos to them.
Google says it's integrating the user end-to-end experience with developer tools to convert engagement into revenue. We'll see how that pans out.
Discovery gets a boost with topic browser pages (for movies, books, games, etc.) and a "Where to watch" feature. Those sound helpful, but I have to use it first to know where I stand regarding it. Audio samples are now available on app store pages, starting with health and wellness apps – an interesting idea.
The Engage SDK is expanding significantly with "Collections" – a full-screen immersive surface to pull content onto users' home screens, continue journeys across devices, create cluster recommendations, and promote specific clusters. Engage SDK content is also coming to the Play Store itself and now supports travel content and reservations. This "Collections" feature actually sounds quite promising for driving deeper engagement, and I'm warm to the prospect of using "Collections" in the future.
Subscription management gets some love too: Single checkout for multiple subscriptions, more visibility for subscription benefits (to fight voluntary churn), and more flexibility for grace periods on payment failures (to combat involuntary churn). The last one particularly stood out for me, especially given sometimes I have to cancel my virtual credit card when info leaks out, so it's nice to have a grace period.
For the game devs, Google Play Games is focusing more on engaging players on Mobile (a shift from PC), an Earnback program to help migrate PC games to Android, the Recall API for cross-platform retention, and bulk achievement creation via CSV. Frankly, I barely play any mobile games, so I don't have much to comment here.
My Sentiment on Google Play: The changes are welcome, but many feel like minor tune-ups rather than any major overhaul. Still, I'm lukewarm towards them. My biggest gripe remains: I feel there’s a lack of serious effort to curb malicious apps. This leads me to hesitate and suspect others when downloading apps from lesser-known developers who haven't shelled out for major marketing. I wish there were an easy way to tell malicious developers apart from genuine ones. Hopefully, AI could help with that, but so far, I've seen no announcement on that topic on Google Play. However, the Engage SDK "Collections" feature does seem genuinely interesting and promising, while the rest seems to be mostly quality-of-life improvements.
Part 3: Android Studio - Narwhal and the Gemini Effect
Android Studio codename "Narwhal" will be getting significant AI adoption, largely thanks to some Gemini powered new features.
- A new KMP template is available, along with Compose preview, menu, and action enhancements. Good.
- Gemini is embedding itself deeper into the IDE: It can help update all project libraries (proposing solutions before touching code), analyze files to create Composable previews, and now it has agent-like capabilities to identify bugs, correct them, and rerun tests. A very promising update.
But the headline act for me is Android Journey, the apparent successor to Android Espresso. The idea of instructing Gemini to test an Android app using natural language commands, rather than wrestling with Espresso's lists, intents, and the dreaded IdlingResources functionality, is a massive improvement. Plus, purging IdlingResource test code from release builds has always been a necessity when using Espresso, a necessity that will become outdated soon. Journey could solve a lot of headaches for exploratory, integration system and regression testing.
Yes, there are potential pitfalls with AI-driven testing, but I believe the benefits here could genuinely outweigh the risks.
Other notable Studio upgrades:
- Firebase device streaming: Test on remote devices via device partner labs. Very welcomed in this age of remote work, for sure!
- ADB Wi-Fi improvements. A nice quality of life improvement, since I've already lost some cables when working with debug apps in Android, hopefully this change will save me some money.
- A new backup feature in Android 16 for app settings and data (presumably with Studio integration). Good.
- Studio will now recommend .ktx syntax over plain .kt. Another minor change.
- Official support for Android XR devices. Good.
- And a heads-up: 16KB page size validation is coming, up from 4KB. Time to check those native libraries. I won't complain about it, in the age of gigabytes of RAM in computers, I believe we are way past the point of needing only 4KB or 8KB memory pages for any partition type or memory context. Still, I think we could have adopted larger memory pages, who knows?
My Sentiment on Android Studio: I’m actually quite excited about the upcoming changes in Android Studio Narwhal. Journey replacing Espresso is a potential game-changer, further allowing Jetpack Compose to assume the role of leading UI designing tool in Android, and helping in the rough edges of many different kinds of testing, especially regression test. The Gemini integration for library updates and bug squashing is also welcomed. Features like 16KB page support and better ADB over Wi-Fi are nice to have, but I see them more as quality of life updates than the real deal for the event.
Part 4: The All-Encompassing Adaptive Android
The push for "Adaptive Android" was a recurring theme, with Google heavily promoting support for various form factors: Phone, Tablet, Foldable, Wearable, Automotive, and the newest product (not so new, actually), Android XR. I believe the dream behind it is that many apps could work seamlessly across different product environments, while retaining the same UI and UX when used, without a huge effort to rewrite those apps. But, in my opinion, there were also another message, an underlying message: If your app isn't adaptive, you might be left out of the whole Android userbase, so you better use it. I think this kind of approach is counterproductive - an approach where a Jetpack Compose library helps the developers to adapt faster, easier and with a minimum cost to their companies, would be much more productive. Let's hope this is the road we are on.
The Navigation 3 library will allow metadata tags for different navigation styles (like list vs. panel) to help with this. We're also getting new animations for app reconfiguration: Levitation (moves a panel from landscape side to portrait center) and Reflow (moves it to the top/bottom in portrait). For Android XR, there are new Android XR exclusive components.
As mentioned earlier, Android 16 will start ignoring orientation, aspect ratio, and resizability restrictions on larger screens (defined as 360dp width or more). Instead of the developer, the user will have total control of the app on larger screens. Hence, it is understandable why Google is advocating for a single, adaptive app for all these form factors, built with Compose and their tooling. And this adaptive approach extends to games too.
My Sentiment on Adaptive Android: While the ambition is clear, and Android was long overdue for an update regarding screen reconfiguration, this heavy emphasis on adaptiveness for larger screen, again feels like a strong-arming Android towards bolstering the app ecosystem for Android XR, even before it is launched. I get the strategy – Android is growing on larger screens, like televisions, totems, monitors, automotive, so a form factor app library becomes necessary for this constantly evolving platform. But the reality for many developers is, they hardly have the time or budget to work on form factors, much less to adapt Android to a supporting platform, especially if it’s not a natural fit for their app. I fear the community and the Android Google team might be on different pages.
Perhaps for games, this one-app-fits-all approach will work wonders, but I suspect many specific apps may not translate well to XR. And I believe Android XR, much like Wear OS, will carve out its own small but dedicated niche, but nothing too far from that.
Regarding using Android XR for something like a translation tool, it may be interesting, but the thought of strapping a $1000+ gadget to my face, ripe for a quick slap-and-grab mugging, still gives me pause. I'm willing to consider using it, and Android XR may be an interesting product, no doubt, but perhaps not the revolutionary leap the presentation painted it as.
That said, the smaller changes aimed at improving Android's behavior across different screen sizes, like the navigation and animation enhancements, are definitely welcome.
So, that's my 2 cents regarding Google I/O '25 track about Android -> A lot of AI, a big push for adaptive everything (especially Android XR, it seems), new libraries and updates, and some genuinely useful improvements sprinkled throughout. Some of it is exciting, some of it feels a bit pushed, and the real test will be when these tools and APIs land in our hands.
I recommend that you take action: Download the newest Android Studio Narwhal, start testing your product with Journey and Gemini. Take some time to consider adapting your app with form factors (consider the time and budget necessary to do so), and don't forget to replace any workaround with a proper, Jetpack Compose based library. Even if it's in alpha development, I think it will be worth it. Time to update those build.gradle files, I suppose!
Top comments (1)
Nice posting, could you share your email address?