Skip to main content

We use cookies to enhance your experience. By continuing to visit this site you agree to our use of cookies. Cookie Policy

AI in ASIA
News

Apple's AI Plan: Gemini Today, Siri Tomorrow?

Apple signs $1 billion deal with Google to power Siri using Gemini AI models, transforming its struggling virtual assistant with advanced language capabilities.

Adrian WatkinsAdrian Watkins4 min read

AI Snapshot

The TL;DR: what matters, fast.

Apple signs $1 billion annual deal to power Siri with Google's Gemini AI models

Partnership uses hybrid approach maintaining Apple's privacy-first infrastructure

New Siri expected in iOS 26.4 with unprecedented personalization capabilities

Apple's Billion-Dollar Deal Transforms Siri with Google's Gemini AI

Apple is making its boldest AI move yet, reportedly signing a $1 billion annual deal to power Siri with Google's Gemini models. The partnership, codenamed "Linwood" internally, marks a dramatic shift for the iPhone maker's long-struggling virtual assistant.

After years of watching Siri fall behind competitors, Apple has chosen Google's 1.2 trillion parameter large language model to breathe new life into its digital assistant. The upgraded Siri is expected to arrive with iOS 26.4 this spring, promising unprecedented personalisation capabilities.

The deal represents more than just a technical upgrade. It signals Apple's willingness to outsource critical AI infrastructure whilst it develops its own proprietary systems, highlighting the intense competition reshaping the tech landscape.

Advertisement

Behind the Scenes: How the Partnership Works

Apple isn't simply handing over Siri to Google. The company will run Gemini models on its own Private Cloud Compute infrastructure, ensuring user data never leaves Apple's ecosystem. For simpler tasks, Siri will continue using Apple's on-device models, whilst complex queries get routed to the Google-powered backend.

"Gemini models and Google cloud tech will underpin Apple Foundation Models for future Apple Intelligence features, including Siri," according to a joint Apple-Google statement from January 2026.

This hybrid approach allows Apple to maintain its privacy-first reputation whilst accessing cutting-edge AI capabilities. Users will experience faster responses, better contextual understanding, and improved integration with personal data across their devices.

The strategic partnership also highlights how even fierce competitors must collaborate in today's AI arms race. Apple's previous exploration of partnerships with Chinese AI companies shows the global nature of this competition.

By The Numbers

  • $1 billion annual cost for Google Gemini integration across 2 billion Apple devices
  • 650 million monthly active users currently use Gemini apps, growing to 750 million by Q4 2025
  • 2 billion monthly users access Gemini AI Overviews, making it the second most-used AI platform
  • 2.4 million active Gemini API users, up 118% from 1.1 million in March 2025
  • $155 billion Google Cloud backlog by Q3 2025, supporting millions of iPhone workloads

Siri's Transformation Timeline

The new Siri represents a complete reimagining of Apple's 2010 virtual assistant. Years of incremental updates failed to close the gap with competitors like Alexa and Google Assistant, leaving Siri as a frequent punchline rather than a productivity tool.

Apple's internal development struggles became evident when key AI personnel departed, including Ke Yang, who led chatbot efforts before joining Meta. The company's Apple Intelligence rollout showed promise but lacked the conversational depth users expected.

Timeline Development Phase Key Features
2010-2020 Original Siri Era Basic voice commands, weather, timers
2020-2025 Apple Intelligence On-device processing, improved accuracy
Spring 2026 Gemini Integration Personalisation, contextual understanding, complex reasoning

The "Linwood" project represents Apple's acknowledgment that catching up requires external expertise. By leveraging Google's proven AI infrastructure, Apple can focus on user experience whilst benefiting from years of machine learning research.

Privacy Concerns and Technical Implementation

Apple faces a delicate balancing act between AI capabilities and privacy commitments. The company promises that sensitive personal data will remain processed by Apple's own Foundation Models, with only anonymised queries reaching Google's systems.

"Google Cloud signed more contracts worth over $1 billion in Q3 2025 than in the combined previous two years," noted Sundar Pichai, highlighting the infrastructure demands of modern AI partnerships.

This technical architecture mirrors approaches used in other Gemini integrations, where local processing handles sensitive operations whilst cloud models manage complex reasoning tasks.

The implementation strategy includes several key privacy safeguards:

  • Personal data processing remains on Apple devices using proprietary models
  • Complex queries use anonymised data sent to Gemini via Apple's cloud infrastructure
  • No direct data sharing between Apple user profiles and Google systems
  • Encrypted communication channels prevent external access during processing

Apple's approach contrasts with competitors who integrate third-party AI services more directly. The company's reputation depends on maintaining user trust whilst delivering competitive AI features.

Market Impact and Industry Response

The Apple-Google partnership sends ripples through the AI industry, demonstrating how even the largest tech companies must collaborate to remain competitive. Microsoft's partnership with OpenAI and Samsung's agentic AI developments show similar strategic alliances shaping the market.

Industry analysts view the deal as validation of Google's AI leadership whilst highlighting Apple's pragmatic approach to innovation. Rather than rushing to market with inferior technology, Apple chose to license proven capabilities whilst developing internal alternatives.

The partnership also raises questions about long-term competitive dynamics. As alternative AI models emerge, Apple may eventually reduce dependence on Google's technology.

Will this partnership affect Apple's relationship with other AI companies?

Apple previously explored partnerships with OpenAI and Anthropic, suggesting the company remains open to multiple AI providers. The Google deal likely represents the best current option rather than an exclusive long-term commitment.

How will this impact Siri's privacy reputation?

Apple maintains that personal data processing stays on-device, with only anonymised queries reaching Google's systems. The company's Private Cloud Compute infrastructure ensures user privacy whilst enabling advanced AI capabilities.

When will users see the new Siri features?

The enhanced Siri is expected to launch with iOS 26.4 this spring, bringing improved personalisation and contextual understanding to supported devices. The rollout will likely be gradual across different regions and device types.

What does this mean for Apple's own AI development?

The Google partnership appears to be a bridge solution whilst Apple develops proprietary AI capabilities. The company continues investing in internal research whilst leveraging external expertise to remain competitive immediately.

How much will this cost Apple annually?

Reports suggest the partnership costs Apple approximately $1 billion per year, reflecting the premium for accessing Google's advanced AI infrastructure across 2 billion Apple devices worldwide.

The AIinASIA View: Apple's Gemini partnership represents pragmatic innovation over pride. Rather than releasing subpar AI features, the company chose proven technology whilst developing internal capabilities. This strategic patience could pay dividends as AI becomes central to user experience. However, the $1 billion annual cost highlights how expensive cutting-edge AI has become. We expect Apple to eventually reduce this dependence through internal development, but for now, users benefit from world-class AI without compromising privacy. The partnership sets a precedent for how tech giants must collaborate in the AI era.

The transformation of Siri from digital assistant to AI companion marks a pivotal moment for Apple's ecosystem. With Google's Gemini providing the intelligence and Apple's infrastructure ensuring privacy, users finally get the smart assistant they've been waiting for.

Will this billion-dollar bet on Google's AI finally make Siri worth talking to? Drop your take in the comments below.

YOUR TAKE

We cover the story. You tell us what it means on the ground.

What did you think?

Adrian Watkins

Adrian Watkins

Founder & Editor

I've spent over 26 years helping companies from global corporations to fast-growing startups achieve measurable success through AI-powered digital transformation, smart go-to-market execution, and sustainable revenue growth. I launched AIinASIA to help share news, tips and tricks for work and play.

Share your thoughts

Join 4 readers in the discussion below

This is a developing story

We're tracking this across Asia-Pacific and may update with new developments, follow-ups and regional context.

Advertisement

Advertisement

This article is part of the AI Tools Power User learning path.

Continue the path →

Latest Comments (4)

Rohan Kumar
Rohan Kumar@rohank
AI
7 December 2025

wow $1 billion for Linwood! Imagine what my agency could do for clients with even a fraction of that budget building out custom Gemini integrations. What kind of ROI do you think Apple is projecting?

Arjun Mehta
Arjun Mehta@arjunm
AI
5 December 2025

yeah, "Linwood" makes sense. abstracting out the LLM integration from client-side inference is actually cleaner for dev teams. good architectural decision.

Daniel Yeo@dyeo
AI
26 November 2025

1.2 trillion parameters and they still can't get Siri to understand my Singlish. Spent enough time debugging NLU models at Shopee to know this is more than just parameter count.

James Clarke@jamesclarke
AI
11 November 2025

It's interesting to see Apple effectively outsourcing a core AI component for Siri, even with the "Linwood" codename. Makes me think about the talent pool we're building up here in Manchester, especially with our university research. Could see some great partnerships form closer to home in the future for similar challenges.

Leave a Comment

Your email will not be published