Joanna Stern spent one week with new Siri AI, and it’s very good

Joanna Stern’s hands-on test of Apple’s new Siri AI in iOS 27 offers more than just a first look at the revamped assistant—it’s a window into how Apple is playing catch-up in a rapidly evolving AI arms race. While competitors like Google and Amazon have already embedded advanced AI into their voice assistants, Apple’s reputation for privacy and seamless integration has long set it apart. The new Siri, with its deeper contextual understanding and more natural responses, signals Apple’s belated but ambitious push to redefine what a digital assistant can do. For users, this shift could mean fewer awkward phrasing commands and more intuitive interactions, but it also raises questions about whether Apple can reconcile its privacy ethos with the data-hungry demands of modern AI. What’s striking here is how Apple’s approach contrasts with the broader industry trend. Most big tech players have leaned into generative AI by relying on cloud-based models that require vast data processing, often raising red flags about user privacy. Apple, by contrast, has hinted at on-device processing for many Siri queries—a nod to its longstanding commitment to limiting data exposure. Yet, as Stern’s testing likely revealed, this constraint may come at a cost: the new Siri may not yet match the depth of cloud-powered rivals, particularly in complex or ambiguous requests. The balance between innovation and privacy is a tightrope Apple must navigate, especially as regulators and consumers grow more sensitive to how their data is used. Looking ahead, the real test will be how quickly Apple can refine Siri without compromising its core principles. The company’s delayed but aggressive AI rollout suggests it sees this as a make-or-break moment—not just for Siri, but for its broader AI strategy. Will users notice the difference? Probably. Will it be enough to close the gap with competitors? That remains to be seen. What’s clear is that Apple’s AI pivot isn’t just about keeping up—it’s about defining how AI should work in a world where trust and utility are increasingly intertwined. Stern’s hands-on experience underscores just how high those stakes have become.