Apple operating systems don’t need an AI year, they need a Service Pack year

It’s now official, then: after the 18-month embarrassment of Apple Intelligence, Tim Cook and his team finally decided to give up trying to build an advanced AI assistant in-house and use Google’s Gemini to power the next iteration of Siri instead. How that plays out we will have to see, of course, but – in the context of the AI hype cycle we are going through – Apple had no choice but to be seen as heavily investing in that area, like every other major tech company in the world does right now. It was inevitable, even only in terms of optics.

But the company’s software troubles are not just about AI. Taking a look back at 2025, it’s plain to see that Apple had an “OK, but could be much better” year overall because many of its hardware products are routinely betrayed by their software. iPadOS, iCloud integration, Liquid Glass, the list goes on. There’s a clear disparity between how most of the company’s products look and what they actually do, between how they are put together and how they work. They still feel as premium as ever, yes, but their software is not of the same quality. It just isn’t.

Consumers do not purchase Apple devices to just enjoy looking at them or holding them in their hands, though: they buy them to use them – and the overall experience is often tainted by software that leaves a lot to be desired nowadays. This has nothing to do with AI or its questionable usefulness. It has everything to do with planning, execution and quality control, which is what Apple needs to focus on in 2026. Here’s why and how.

The problem: too much attention on marketing, too much AI hype

It’s not that hard to describe how this imbalance between Apple’s hardware and software came to be: these two sides of the company’s business have simply been operating based on different principles. While Apple’s hardware teams have carefully but steadily iterated on the architecture of the M- and A- series of chips, based on roadmaps and specific ways those can evolve, Apple’s software teams seem to not have a clear vision and a specific direction they want to take with the company’s operating systems as a whole.

As a direct result of this, the last three versions of iOS, macOS, iPadOS and watchOS have suffered from a severe case of featuritis: the attempt to add new features that do not serve a specific purpose or pave the way to a well-defined future – sometimes even announced seemingly for the sake of it – making them look more like marketing bullet points than meaningful operating system advances.

At the same time, a lot of attention was given to unnecessary or pointless AI functionality while not enough attention has been given to quality control and code testing, areas where Apple used to put a lot more effort into. Since this is Apple we are talking about here, this was/is not a matter of resources, but rather a company’s conscious choice to simply not pursue software excellence for its hardware platforms anymore.

One only has to look at Liquid Glass – Apple’s new graphical user interface – in order to understand exactly what the problem is with the company’s software approach over the past few years. The actual purpose of Liquid Glass is still not clear to anyone, it does not make using Apple’s devices any easier (many will attest to the contrary), it has a negative effect on operation speed and battery life, it shipped in an unfinished state and it’s being fixed while being used on hundreds of millions of consumer devices.

Needless to say, this is not what Apple fans had come to expect and used to love about the company’s particular, but effective and once-polished user experience. It’s the exact opposite of it.

The solution: a year of focus on code quality instead of pointless AI

It’s obviously important to note that Apple’s operating systems are still user-friendly, still effective, still secure and private compared e.g. to Windows or Android. But if Apple wants to avoid putting that “it just works” reputation of its products on the line, it needs to get back to basics: to focus on the code quality of its operating systems, on the consistency of their user interfaces, on the effectiveness of their functionality.

Truth be told, Apple’s current operating systems do not necessarily need more new features – but they definitely need the kind of polishing that comes from extensive testing, relentless bug fixing and careful optimization.

Apple fans may not remember this, but Windows users – especially long-time ones – certainly do: once upon a time, before Windows 8, Microsoft was in the habit of doing “Service Pack” releases of its operating systems. It would release, for instance, Windows XP or Windows Vista or Windows 7 and, one year in or two years in or even three years in, it would release a “Service Pack” update instead of a new version.

The interesting thing was that those Service Packs did not just contain all the security patches or fixes Microsoft offered during the year before: they also contained optimizations and fixes for various other things, sometimes undocumented, that resulted in a more stable, more reliable, occasionally more performant a version of Windows. Same OS, but the SP1 or SP2 or SP3 version of it, just sporting tighter, better code than before.

This is what Apple operating systems truly need in 2026. They need a Service Pack year, not an AI year. Yours truly does get the peer pressure when it comes to what everyone else in the tech world seems currently to be doing, but if there’s one company that has the luxury to ignore it, it’s this one. Hardly anyone is asking for “more AI” in all Apple operating systems, but almost everyone is asking for this or that or the other OS issue to be addressed – and it’s the executives’ job to illustrate, through their choices, that what consumers out there are insistently asking for is more important than what investors happen to be currently hyped about.

What the company should do, then, is release the Gemini-driven Siri whenever it’s actually ready and, for this year, focus on what truly matters when it comes to operating systems: stability, speed, security, GUI uniformity, dependable functionality, exemplary resource handling, optimal power management. See to it that this is a Service Pack year for all Apple operating systems, even if that means no fancy new features for 2026.

At the end of the day, people do not enjoy using their devices based on feature lists or marketing bullet points. They love or hate any device’s user experience based on said device’s OS fundamentals. Here’s hope that Cook and his team, amid all the AI craze of recent years, have not lost sight of that.