Apple’s decision to rebuild Siri around models developed with Google marks a pivotal shift in how the company approaches artificial intelligence. Rather than relying solely on its internal Apple Foundation Models, the company has integrated Gemini-derived capabilities to deliver context-aware assistance that can reference apps, conversations, and web data without requiring users to switch contexts. This move arrives alongside aggressive hardware support cuts and a deliberate emphasis on focused software experiences, revealing a strategy that prioritizes AI depth over broad device longevity.
The partnership underscores Apple’s willingness to blend its privacy architecture with external model performance. Five new foundation models—AFM 3 Core, AFM 3 Core Advanced, AFM 3 Cloud, AFM 3 Cloud (Image), and AFM 3 Cloud Pro—form the technical backbone, with the 20-billion-parameter multimodal model positioned as the most capable on-device option. These systems power the revamped Siri AI, which will appear across iPhone, Mac, Watch, and AirPods once the beta arrives later this year.
Siri AI Gains Context Through Google Collaboration
Craig Federighi emphasized that Apple is not simply rebranding Gemini. The company uses none of Google’s customer-facing models or search infrastructure for grounding knowledge, instead routing queries through its Private Cloud Compute system that promises no retention of user data or chat logs. This hybrid approach allows Siri to handle complex, multi-step requests—such as retrieving a podcast mentioned in a text conversation or resuming prior interactions across devices—while maintaining the on-device and private-cloud guarantees Apple has promoted since the initial Apple Intelligence announcements.
The technical implications extend beyond the assistant itself. Developers can now build against Core AI frameworks that leverage these same models, potentially accelerating third-party features that require reasoning or image understanding. Yet the reliance on Google for foundational model performance also introduces questions about long-term differentiation, especially as competitors integrate their own large language models more deeply into hardware.
WatchOS 27 Enforces the Sharpest Support Cut in Years
The most immediate consequence for existing users appears in watchOS 27, which drops five models that received watchOS 26: Series 6, Series 7, Series 8, the first-generation Ultra, and the second-generation SE. Only devices with the S9 or S10 chip remain eligible, representing the largest single-year reduction in Apple Watch compatibility.
This cull aligns with similar, though less severe, reductions across other platforms. macOS Golden Gate ends support for the final Intel Macs, while iPadOS 27 raises the floor to A14 or M1 silicon, eliminating five iPad models still running the prior release. The pattern suggests Apple is accelerating the pace at which older hardware exits the update cycle to reduce engineering overhead for AI features that demand newer neural engines and memory configurations.
For owners of the affected watches, the decision creates a clear fork: remain on watchOS 26 or purchase a Series 9 or newer device. The absence of compelling new watchOS 27 features beyond the Siri upgrade has already prompted some users to delay upgrades, highlighting the tension between Apple’s hardware refresh cadence and the value proposition of its software updates.
Focused Experiences Emerge Amid AI Complexity
Alongside the AI-heavy narrative, Apple introduced a deliberately minimalist Sports app that performs only one function: displaying live scores, standings, and schedules. The app avoids news feeds, streaming guides, and algorithmic recommendations, offering instead a clean interface with an optional toggle to remove betting odds. Its arrival during the World Cup period underscores a recognition that many users simply want reliable, low-friction access to game information without the feature bloat that has accumulated in league-specific and media apps.
This philosophy extends to broader iOS 27 and iPadOS 27 updates, which prioritize stability and incremental quality-of-life refinements over dramatic redesigns. The emphasis on Siri AI as the flagship addition reflects a calculated bet that contextual assistance will deliver more everyday value than visual overhauls or additional productivity tools.
Hardware Teases Signal Foldables and Touch Macs
Beta code and new APIs within iOS 27 and macOS Golden Gate contain multiple references to upcoming hardware. Landscape orientation handling, dynamic display count detection, and enhanced resizing behaviors in iPhone Mirroring all point toward a foldable iPhone Ultra expected in September. Similar clues—touch-friendly gestures, expanded markup tools, and adjusted UI spacing—suggest the first touchscreen MacBook Ultra will follow the same timeline.
A privacy-focused HomeKit Secure Video camera is also under development, with the updated Home app adding 4K support and facial recognition features that align with the rumored device’s sensor capabilities. These leaks through software updates illustrate how Apple’s historically tight control over product information has loosened under the pressure of preparing developers for new form factors.
App Store Guidelines Target Low-Effort Submissions
Updated App Review Guidelines now warn that well-established categories—wallpaper apps, simple timers, sound effects, and others—may face removal if they are not meaningfully updated or attracting users. The change formalizes a long-standing “fart app rule” while giving developers explicit notice through the existing App Store Improvements process. Apple frames the policy as a way to reduce clutter and improve discovery, complementing new personalized recommendation features introduced at WWDC.
The move carries business implications for smaller developers who have relied on low-maintenance apps in saturated categories. Repeated submissions of low-quality experiences could now result in removal from the Apple Developer Program, raising the bar for participation in the ecosystem.
These developments collectively illustrate Apple’s effort to modernize its software foundation around capable AI while tightening hardware requirements and curating the App Store more aggressively. The coming months will reveal whether the Siri AI experience justifies the support cuts for users and whether the Google partnership ultimately strengthens or complicates Apple’s claims of end-to-end control.