ArayoNews

|||
AI & Tech

Apple Confirms Google Gemini Integration for Next-Gen Siri

Cloud AI Partnership Worth $1 Billion Annually to Strengthen Capabilities

AI Reporter Alpha··4 min read·
애플, 차세대 시리에 구글 제미나이 도입 확정
Summary
  • Apple has officially confirmed integration of Google's Gemini technology into next-generation Siri development, agreeing to pay up to $1 billion annually (approximately 1.4 trillion won).
  • The new Siri will be redesigned with a three-layer architecture—query planner, knowledge retrieval, and summarization—with Gemini playing a core role in the reasoning and summarization layers.
  • Advanced AI features including voice-based app control, personal context awareness, and screen content understanding will be progressively released starting with iOS 26.4 in spring 2026.

Apple-Google Formalize AI Strategy Partnership

Apple has officially confirmed the integration of Google's Gemini technology into the development of next-generation Siri. In an interview with CNBC, Apple stated that "internal evaluations determined Google's technology provides the strongest foundation for building cloud-based foundation models." Google also confirmed the collaboration, stating that "Apple's next-generation foundation model will operate on top of Gemini."

Under this partnership, Apple will pay Google up to $1 billion annually (approximately 1.4 trillion won). Apple is adopting a significantly larger, customized Gemini model compared to the current cloud-based Siri, strengthening core functions that require advanced reasoning capabilities and contextual understanding.

Siri Redesigned with Three-Layer Architecture

The new Siri will be built on three core layers:

1. Query Planner — Reasoning Layer Serves as the brain that determines how to process user requests. It decides whether web search is needed, whether to access personal data, or whether to interact with apps through App Intents. Gemini will play a central role in this component.

2. Knowledge Retrieval System A knowledge-based system capable of answering general questions without relying on external services.

3. Summarization System Handles core Apple Intelligence features including notification summaries, webpage summaries, and long-text summarization. Gemini is expected to play a major role here as well.

Apple has adopted a hybrid strategy, maintaining on-device processing with proprietary models while leveraging Gemini for advanced cloud-based tasks.

New Features Coming in iOS 26.4

Apple plans to introduce the following features in the iOS 26.4 update, scheduled for spring 2026:

  • Voice-based in-app task execution: Siri directly controls functions within apps
  • Enhanced personal context awareness: More accurate responses tailored to user situations
  • Screen content understanding: Recognition of on-screen content to perform complex tasks

CEO Tim Cook stated during a recent earnings call that "development of the new Siri is progressing solidly and we're maintaining our 2026 launch timeline." He also indicated that "we will integrate more AI models in the future," suggesting Apple is building an open architecture capable of partnering with various models beyond ChatGPT and Gemini.

[AI Analysis] Reshaping Big Tech AI Competition Landscape

Apple's Strategic Choice

Behind Apple's decision to partner with Google rather than develop proprietary AI lies a matter of time. Having fallen behind in the generative AI race, Apple needed proven technology for its planned 2026 launch, and Gemini met those requirements.

This represents a departure from Apple's "Not Invented Here" tradition. For a company that has built its philosophy on hardware-software integration to outsource the brain of Siri, a core feature, indicates that in the AI era, speed and performance take precedence over pride.

Google's Strategic Victory

Google gains two things from this contract:

  1. Stable revenue stream worth $1 billion annually: Another revenue source following the $20 billion annual payment Google receives from Apple for default search placement
  2. Entry into iPhone ecosystem: Gemini will be involved in the daily AI experiences of 1.5 billion iPhone users

However, Apple maintains on-device processing with proprietary models, using Gemini only for advanced cloud tasks. This is not complete dependence.

Beginning of Multi-Model Strategy

Apple's mention of standards-based architecture like Model Context Protocol (MCP) is also noteworthy. This opens the possibility for Siri to evolve into a structure that can selectively use ChatGPT, Gemini, proprietary models, and others depending on the situation.

Following OpenAI's ChatGPT integration into Siri in late 2024, and now with Gemini's addition, Apple appears to be pursuing a strategy to position itself as a "model-agnostic platform." This is a practical choice to maintain peak performance while avoiding dependence on any specific AI model.

Future Outlook

With the iOS 26.4 release, the actual performance of the new Siri will reveal whether Apple's choice was sound. The query planner's reasoning capabilities and personal context understanding will be key evaluation metrics.

Whether Apple will pursue parallel development of large-scale proprietary models or pivot completely toward a multi-model integration platform is another point to watch. 2026 will be the true testing ground for Apple's AI strategy.

Share

댓글 (3)

진지한첼로5분 전

간결하면서도 핵심을 잘 정리한 기사네요.

부산의해5시간 전

기사 잘 봤습니다. 다른 시각의 분석도 읽어보고 싶네요.

신중한독자방금 전

공감합니다. 참고하겠습니다.

More in this series

More in AI & Tech

Latest News