Apple unveils “Siri AI” at WWDC, but uneven rollout and beta timing raise questions

Apple kicked off WWDC 2026 on Monday with a major pitch for generative AI in everyday computing: “Siri AI,” a reworked Siri designed to be more conversational and more integrated across iPhone, iPad, Mac, and other Apple devices. The company said developers can begin testing the new assistant immediately, with a wider consumer beta planned for later this year. But the launch also surfaced immediate red flags, including a region-limited release tied to Europe’s Digital Markets Act and the ongoing reality that Siri AI is still being tested rather than broadly shipped.

In its announcements, Apple framed Siri AI as an “entirely new version of Siri,” powered by Apple Intelligence. The assistant is intended to combine on-screen awareness, personal context, and world knowledge, so it can answer questions, surface information from messages, emails, and photos, and respond to what’s happening on the user’s display. Apple also said Siri AI includes a dedicated app to revisit conversations across devices, expanded “Visual Intelligence” across iPad and Mac (and integrated into Camera on iPhone), and new writing tools meant to generate and edit text directly where users type.

Apple positioned the rework as both a capability upgrade and a privacy architecture shift. In Apple’s description, Siri AI leverages on-device processing plus “Private Cloud Compute,” with the company saying that when Private Cloud Compute handles requests, a user’s personal data is not stored or made accessible to Apple or others. Apple also said outside experts can verify its privacy promise. The company said users can access “frontier-level intelligence” while keeping sensitive information protected, and it added that systemwide actions depend on an on-device orchestrator tied to features like the Spotlight index.

Regulatory friction, however, quickly moved from background noise to a central limitation. In a separate Apple Newsroom update tied to the DMA, the company said Siri AI cannot be shipped in the European Union on iPhone or iPad when iOS 27 and iPadOS 27 launch later this year. Apple said EU users will instead get Siri AI on macOS 27, visionOS 27, and watchOS 27. Apple blamed EU regulators for rejecting proposed solutions that it said would preserve privacy and security, and it said it currently has no timeline for getting Siri AI to EU iPhone and iPad. It also said developers in the EU will not be able to test or use Siri AI features on iOS and iPadOS at that time.

Beyond Europe, Apple also outlined constraints that could shape user perception during the beta period. Siri AI’s consumer availability is described as a beta later this year for users with a supported device set to English, with Apple saying it will expand support for more languages. Separately, in its broader Apple Intelligence update, Apple said some AI features that rely on powerful server models have daily usage limits, and it linked increased access to most iCloud+ subscription plans. Together, those conditions mean Siri AI may roll out in phases, with performance and availability varying across users and use cases.

The caution here is not just about whether Siri AI can answer better questions. It is also about whether Apple can sustain an AI experience that works reliably enough, safely enough, and consistently enough across jurisdictions to persuade users to treat Siri as a daily assistant rather than a novelty. TechCrunch described the revamp as Apple finally bringing a promised generative Siri upgrade after earlier delays, and Axios said Apple has fallen behind in AI amid multiple delays. What to watch next is whether Apple can deliver a smooth public beta, how quickly it expands beyond English, and, critically, whether the EU dispute ends in a time-bound path for iPhone and iPad users to get Siri AI on day one. Apple did not provide a date for EU iOS and iPadOS availability in its DMA update, leaving a key uncertainty hanging over the rollout.