Apple's revamped Siri could tap Google Gemini for AI features, web search

Apple is reportedly developing a new version of Siri powered in part by Google's Gemini AI, adding advanced search, contextual understanding, and deeper integration with Apple Intelligence

Apple’s upcoming version of Siri could be powered in part by Google’s Gemini AI, according to a new report by Bloomberg’s Mark Gurman. The report claims that the company’s upgraded voice assistant “will lean on Google’s Gemini model” and gain new features such as AI-powered web search.

 

This development comes shortly after Apple CEO Tim Cook confirmed in an interview with CNBC that the company plans to expand its Apple Intelligence suite through partnerships with more third-party AI providers, beyond its existing collaboration with OpenAI’s ChatGPT.

Apple’s two-track Siri development

Gurman previously reported that Apple has been testing two separate versions of the new Siri — one powered by Apple’s own on-device AI models and another supported by Google’s Gemini through Apple’s Private Cloud Compute system. However, Apple’s internal version reportedly still falls short of expectations, even after over a year of additional development.

 

Apple has been under growing pressure to catch up in the generative AI space, lagging behind competitors like Google and Samsung. At WWDC 2024, the company announced Apple Intelligence — its on-device AI suite offering tools such as Genmoji, AI-assisted writing, and ChatGPT integration. Yet the most anticipated feature — a re-engineered Siri capable of understanding personal context, processing on-screen information, and executing in-app commands — was delayed.

 

Originally expected to roll out during the iOS 18 cycle (with updates such as iOS 18.4 or 18.5), the upgraded Siri was postponed as Apple opted to rebuild the assistant to meet its quality standards. Senior Vice President of Software Engineering Craig Federighi later confirmed that more time was needed to ensure the new Siri delivered a consistent and reliable experience.

Siri’s Gemini foundation and architecture

Bloomberg’s report suggests that Apple’s partnership with Google may help overcome some of the challenges it faced developing its own large-scale models. Google already offers advanced assistant features through Gemini on Android — including contextual awareness, multitasking, and integration with third-party apps like WhatsApp and Spotify.

The new Siri will reportedly consist of three core components: a query planner, a knowledge search system, and a summariser. Google’s Gemini models are expected to handle the planner and summariser roles, generating structured responses and summaries for user queries.

 

These processes will run on Apple’s Private Cloud Compute infrastructure to maintain data privacy and prevent external data sharing.

 

According to 9to5Mac, citing Bloomberg’s follow-up report, even the knowledge search layer could use Gemini’s models. This would give Siri broader “world knowledge” for answering general questions, expanding beyond the assistant’s current reliance on ChatGPT for complex or contextual requests. In its existing setup, Siri can forward user queries to ChatGPT with permission, but the Gemini-backed integration is said to be deeper.

How Apple may position Gemini-powered Siri

9to5Mac also reported that Apple may not explicitly promote its collaboration with Google. Instead, Gemini’s technology would quietly power parts of Siri, while Apple would market the assistant as an AI experience running securely on Apple servers.

 

This approach would mirror how Samsung integrates Google’s Gemini features into its own Galaxy AI suite — where many of the AI functions are built atop Google’s models but wrapped in Samsung’s interface and branding.