Apple Ships 450M AI-Capable iPhones, Siri Update Looms

Counterpoint Research data, cited in trade coverage, shows Apple has cumulatively shipped 450 million iPhones capable of running generative AI features through Q1 2026. Apple's Apple Intelligence product page lists on-device features such as writing tools, Live Translation, voicemail summarization, and the Image Playground image-generation app. Ahead of its WWDC 2026 keynote on Monday, June 8, Bloomberg and TechCrunch reported that Apple would present a major Siri overhaul and additional Apple Intelligence updates, with a chatbot-style Siri, broader contextual understanding, and an AI agent integration for the App Store. The new software is reported as iOS 27.

The Siri revamp builds on a partnership Apple and Google announced on January 12, 2026, under which the next generation of Apple Foundation Models will be based on Google's Gemini models and cloud technology. The companies' joint statement said, "Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple's industry-leading privacy standards." Reporting from Bloomberg and TechCrunch indicates iOS 27 may also let users choose third-party models for Siri.

As a generic industry pattern, devices with dedicated NPUs and larger memory are necessary but not sufficient for broad generative experiences; vendors typically pair smaller on-device models with cloud-hosted large models to balance latency, privacy, and capability. For practitioners, cross-device features usually require hybrid architectures, local inference for private data and selective cloud calls for heavy reasoning, plus careful handling of model fallbacks and consistency between on-device and server outputs. A 450-million-device base materially lowers distribution friction for new features compared with a greenfield deployment, though The Verge and others note Apple's earlier Apple Intelligence rollout missed some promised capabilities.

• •Whether Apple ships system APIs letting third-party developers plug into the new Siri and chat interfaces.
• •Clear on-device versus Private Cloud Compute execution paths for heavy models.
• •SDKs or agent tooling exposing multi-step task flows, where most developer impact will appear.