If you take a shot every time someone mentions “agent” AI at the latest tech keynotes — Google I/O, Microsoft Build, and NVIDIA’s Computex blowout — you’ll get sick in no time. This is the latest industry term to describe AI agents that can do work on your behalf without any direct input, like automatically adding a meeting to your calendar based on your email. It’s as if the tech world can’t wait to sit back and let AI take over. We’ll probably see impressive agentic AI within a decade, but I’m worried about jumping into the world of agents with our current batch of AI models, which may still be confusing and not entirely reliable. The idea of letting current AI agents function completely on their own seems like sheer madness.
Apple is once again thinking something different. Agentic AI was only briefly mentioned during the WWDC 2026 keynote this week. Instead, Apple spent most of the time talking about how its new Siri AI could actually be useful: like finding a friend’s new address hidden in a long text message thread, or figuring out how to get tickets to a particular concert. For the most part, Siri AI responds to your commands, it has the advantage of modern AI models to better synthesize the data.
Based on our early look at Siri AI at WWDC 2026, it appears to work as advertised, although we’ll need to do long-term testing to see if it really holds up. We’re also dealing with the first developer beta, so there’s plenty of room for things to change between now and when Apple’s latest OS updates arrive this fall. But as an AI skeptic, I’m surprised to find myself more interested in the possibilities of Siri AI than I am in anything approaching Microsoft’s CoPilot.
Apple’s commitment to privacy with private cloud compute also gives it an edge over Google and Microsoft — the company says it only uploads relevant data, makes it all anonymous and doesn’t track server logs. For its new AFM3 Cloud Pro model, which runs on Google’s servers with NVIDIA-powered hardware, Apple has also upgraded private cloud compute to provide the same amount of security.
“While we absolutely reduce the amount of content sent to the PCC, the important thing about the PCC is that architecturally, it’s an efficiency measure at that point,” Craig Federighi, Apple’s SVP of software engineering, said during a WWDC panel. Ars Technica). “Because the PCC itself, by design from the beginning, will evaporate any record of that data immediately after it answers your question… it’s not stored. It’s all in a form where it’s completely ephemeral.”
However, Apple couldn’t completely resist the allure of agentic AI. And not surprisingly, it’s also linked to one of the more troubling new Apple Intelligence features: the ability to automatically change your compromised passwords. Apple claims the new Password app helps people “securely navigate through websites to sign in and upgrade their accounts to stronger passwords.”
The potential issues are obvious: What else might happen when Apple Intelligence logs into that website? Can I really trust the password that comes with it? Then again, I also know that I still have compromised passwords on unimportant websites, wouldn’t it be better to recover them automatically? The goal of making users more secure may be worth a slight loss of agency through Apple’s agent, but it’s a slippery slope.
Apple is also using agentic AI with Safari’s new “Notify Me” feature, which lets you set alerts for specific changes to a website. This can be useful for tracking price changes or news updates around a specific organization – it’s certainly much better than keeping a tab always open and constantly refreshing it. It’s easy to see how Apple could take this feature further: Maybe eventually AI could automatically purchase a hot product before it’s out of stock, or send an email to a group of friends if you’re keeping an eye on vacation property pricing.
You could argue that Apple is far behind its competitors when it comes to AI in general, and you’d have a point. Microsoft and Google rushed to put their AI capabilities into every corner of their consumer products, making their investors happy (even if it made those products worse). Meanwhile, Apple has struggled with delays for its AI-powered Siri upgrade, which was first announced two years ago. The first batch of Apple Intelligence features, like Notification Summaries and Zenmojis, were pretty weak (and in the case of those summaries, it made some embarrassing mistakes).
However, Apple is rarely the first in new technology. It didn’t create the first MP3 player, smartphone or ultraportable laptop. Instead, the company has generally focused on fixing serious issues of early market entrants, making its solutions more useful to consumers, and, of course, charging a ton of money for the privilege of using its more sophisticated products. The iPod was successful because it could hold a lot of music, was easy to use, and was linked to iTunes, the first major attempt to make digital music legally available. The iPhone succeeded by surpassing keyboard-bound smartphones with a larger screen and more useful apps.
With Siri AI, Apple has the vision to see where Microsoft failed with Copilot, and it has a clear vision of the AI features people will actually want to use. As its models improve and consumers become more accustomed to AI, it will undoubtedly lean toward more agentic features. I’m just hoping that Apple continues to be judicious about how it implements agentive features. I’m already worried about Siri mishearing my voice commands, I don’t want to worry about it draining my bank account.
<a href