Apple Picks Google to Build Next-Gen Siri on Gemini

LAS VEGAS: Google has publicly confirmed that Gemini will power a more personalised version of Siri, arriving later this year. Google Cloud CEO Thomas Kurian made the announcement at Google Cloud Next 2026 in Las Vegas. The confirmation adds new detail to a partnership Apple and Google announced jointly in January.

Apple and Google struck a multi-year deal at the start of 2026 under which the next generation of Apple Foundation Models would be built on Gemini technology and run on Google Cloud infrastructure. Apple evaluated offerings from OpenAI and Anthropic before choosing Google.

The deal is reported to cost Apple approximately $1 billion per year. Kurian confirmed that Google is Apple’s “preferred cloud provider”, the same language used in January, now applied specifically to Siri.

The upgraded Siri is expected to debut at Apple’s WWDC developer conference on June 8 and ship with iOS 27 in the autumn. Apple had originally targeted a spring 2026 launch but adjusted its timeline after running into accuracy issues during internal testing.

The Gemini-powered features are expected to make Siri genuinely conversational, capable of chaining queries, retaining context, and taking actions across apps rather than answering isolated voice commands.

Privacy mechanics remain unresolved publicly. Apple built its existing AI infrastructure around on-device processing and Private Cloud Compute, keeping data off third-party servers. How much Gemini processing will run on Google’s infrastructure versus Apple’s own remains unclear, and unconfirmed by either company.

Apple has reportedly asked Google to prepare dedicated server capacity to handle the increased cloud demand the smarter Siri will generate. Both companies say Apple’s privacy standards will be maintained throughout the partnership.