Google Confirms Gemini Will Power Apple's New Siri: A Two-Phase Rollout Through 2026

The Announcement

At Google Cloud Next '26 in Las Vegas on April 22, 2026, Google Cloud CEO Thomas Kurian made a landmark announcement: Google and Apple are formally collaborating on next-generation Apple Foundation Models built on Gemini technology. These models will power future Apple Intelligence features — including a deeply reimagined version of Siri expected to ship later this year.

Kurian stated: "We're collaborating with Apple as their preferred cloud provider to develop the next generation of Apple Foundation Models based on Gemini technology."

This marks the first time Google has publicly confirmed the scope of the Apple partnership at a major conference, elevating it from rumor to official roadmap.

What Changed and Why It Matters

Apple's Siri has struggled to keep pace with AI assistants from Google, OpenAI, and Anthropic. Plans to ship advanced Siri capabilities with iOS 18 were famously delayed. Apple's own foundation models, while optimized for on-device efficiency, have not matched the reasoning depth of frontier cloud models.

By selecting Google Cloud as its preferred cloud provider and grounding its next foundation models in Gemini technology, Apple is effectively outsourcing the heavy-lifting of AI reasoning to one of the world's most capable LLM families — while retaining control over the user-facing experience and on-device privacy architecture.

Two-Phase Rollout

The partnership is executing in two distinct phases:

Phase 1 — iOS 26.4 (Spring 2026, Already Underway)

Siri has begun using Gemini for enhanced context awareness and on-screen recognition. Users on iOS 26.4 can experience improved follow-up query handling and more accurate responses that draw on what is visible on their screen. This phase operates transparently — users do not need to take any action.

Phase 2 — iOS 27 (Fall 2026, alongside iPhone 18)

The full "Conversational Siri" experience launches with iOS 27, expected in September 2026. This version introduces:

• A standalone Siri app with persistent chat logs
• A chatbot-style interface for extended conversations
• Multi-action parsing: Siri will execute multiple sequential tasks from a single natural-language command
• Deeper integration with third-party apps via expanded SiriKit capabilities

Apple is expected to reveal the complete feature set at WWDC 2026 on June 8.

Infrastructure: Apple on Google Cloud

A key technical element of the partnership is infrastructure. Apple has engaged Google to provision server capacity within Google data centers to run Siri's cloud-based inference workloads. This is significant because Siri's scale — hundreds of millions of active Apple devices — demands compute infrastructure few organizations can match.

Apple's selection of Google Cloud over AWS or Microsoft Azure is also strategically notable. It signals that Apple views Gemini's model quality and Google Cloud's reliability as the optimal foundation for its AI ambitions.

Competitive Implications

The announcement has immediate implications for the competitive landscape:

For Google: Gemini technology now runs on Apple hardware for hundreds of millions of users — a massive distribution advantage. Revenue from Apple Cloud commitments strengthens Google Cloud's enterprise position.

For OpenAI and Anthropic: Apple's choice of Gemini over OpenAI's models (which already power ChatGPT integration in iOS) or Anthropic's Claude narrows the path for those companies to become Siri's primary intelligence layer.

For consumers: iPhone users gain access to reasoning-grade AI through Siri without needing to switch to a dedicated third-party AI app — the implications for daily workflow automation are substantial.

The Personalization Layer

Apple has consistently emphasized privacy as a differentiator. The Gemini-powered Siri maintains Apple's privacy architecture: on-device processing handles sensitive data locally, while cloud inference through Google's infrastructure handles complex reasoning tasks under Apple's data agreements. Apple's Foundation Models serve as an intermediary layer, shaping how Gemini's outputs are filtered, personalized, and presented to users.

Outlook

The Google-Apple collaboration represents one of the most consequential AI partnerships announced in 2026. If the Phase 2 Siri delivers on its promise, Apple could reclaim relevance in the AI assistant space with a product that combines Google's frontier reasoning with Apple's device integration and privacy reputation. WWDC 2026 in June will be the first full public preview of what this partnership produces in practice.