Apple's new Siri just works well. That's enough

There's an old saying in the product world: you don't have to be the first, just the one who gets it right. Apple has lived that philosophy with the iPod, with the App Store, and now in quite an obvious way, with AI. WWDC 2026 is nothing particularly explosive. The new Siri does not have a complex automated AI agent, no epic robot demo that does everything for you. What Apple showed was AI at work: quietly, consistently, and helpfully, across their entire ecosystem. And when compared to what Microsoft is doing with Copilot, that contrast speaks volumes.

Apple is not early, not late, they like to do it just in time

It would be unfair to say Apple actively waited. The reality is simpler: they are slow, afraid of change, gossiping about WWDC 25 and Apple Intelligence. And the proof is that Apple is still using Google Gemini for part of its AI services instead of doing it completely on its own. That's not a sign of a company that's been ready for a long time, but they still have their problems.

WWDC26 is where Apple tried to right things wrong with Siri AI

One Siri, everywhere

I read some reviews, and there is one point I like that was confirmed by PCWorld's Mark Hachman, which is also a very sharp comment: Apple has one Siri, they only need one, while Microsoft has many. Specifically: if you use Windows and want to search smart files, you have Windows Search, or File Explorer with semantic search added a few months ago, or Command Palette in PowerToys, or Copilot app. But Copilot app cannot find local files. File Explorer can search, but most users don't know that feature exists. Command Palette finds files but has no intelligence. If you right-click on a file in File Explorer, you see the “Ask Copilot” option, but only for a single file.

iOS 27, macOS 27 is how Apple tries to get AI right And in fact, if the system is designed with a consistent mindset, users will not have to ask why and what to do. The results Microsoft has achieved are the result of many teams, many times, many independent decisions, and users have to find a way to put them together. Meanwhile, Spotlight on macOS will now integrate Siri directly. You ask a vague question, Siri understands the context, digs through emails, messages, files, and returns results. No need to know which app to use. No need to remember which feature was added last month.

Lessons from Windows Recall

Windows Recall is a great feature, but Microsoft doesn't do it right But Microsoft's initial implementation did not sufficiently encrypt that screenshot data, leading to a fierce wave of "spyware" criticism. Microsoft must constantly update and explain, but lost trust is very difficult to regain. Apple does the same thing: save context, search throughout the application, but with a commitment to processing on the device, not sending data to the cloud. Result: same behavior, but completely different user reaction. One side is considered a useful feature; one party is considered a privacy threat. This is not purely a technical issue but has become an issue of branding and historical relations with users.

Local AI as a competitive advantage

One of the points Apple emphasized at WWDC this year is that most AI Intelligence tasks run on-device: on the device's chip, not through a server. For more complex tasks, Private Cloud Compute is needed, and Apple commits not to store user data during that process.

On-device AI is how Apple promotes their AI, and it is a huge competitive advantage What's interesting is that users don't necessarily need to understand the technical details to feel the difference. They just need to know: their data isn't going anywhere. And in the context of increasing privacy concerns, the feeling of "AI running right on your machine, no one sees" is a marketing advantage that Microsoft is having a hard time creating, not because it doesn't have the technical ability, but because the Windows ecosystem is too fragmented to make the same consistent commitment.

Of course, Apple still doesn't have complete control

It would be unfair to just praise. Apple Intelligence is still relying on Google Gemini for some tasks. That means not all AI experiences are under Apple's control. And more sophisticated AI agent features, the kind that can automate multi-step workflows, working across applications without user intervention, have yet to appear at the level that competitors like OpenAI or Google are deploying. Apple is making a steady move rather than a bold move. The question is will that approach still be enough as competitors move faster in pushing the limits of AI, or will the gap start to shrink in Apple's favor? Temporarily, in 2026, the answer is still in favor of Apple. Not because they have the best AI, but because they are the only ones who make the AI ​​feel like a natural part of the device, instead of an added feature.