Apple hopes to save Siri from laughingstock status with infusion of Google Gemini

Partnership between behemoths raises questions about OpenAI's place at the iTable

It may finally be time to take AI on the iPhone siri-ously. Apple and Google on Monday announced a multi-year partnership that will see Apple Foundation Models standing on the shoulders of Google Gemini models, one that will return a small portion of the roughly $20 billion Google pays annually to be Apple's default search provider.

Terms of the tie-up have not been disclosed, but Bloomberg previously reported that Apple was planning to pay about $1 billion per year to utilize Google's AI technology.

"After careful evaluation, Apple determined that Google's AI technology provides the most capable foundation for Apple Foundation Models and is excited about the innovative new experiences it will unlock for Apple users," the two companies said in a joint statement. "Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple's industry-leading privacy standards."

Apple's home-grown machine learning, despite some effort to show off rather than hide behind the usual secrecy, has been beset by unkept promises and staff departures. An updated Gemini-flavored version of Apple Intelligence is expected to debut later this year.

The deal looks like a setback for OpenAI, brought in to buttress Siri in 2024 with support for ChatGPT as a Siri extension. Last month, OpenAI CEO Sam Altman reportedly declared a "Code Red" to signal the urgency of responding to improvements in Google's Gemini model family.

Apple, in theory, has some strong cards to play in the AI game, specifically hardware suited for running local models and its public commitment to privacy, which like security has been something of an afterthought among commercial AI model makers. 

Apple's Private Cloud Compute system for handling AI queries in a way that protects private data proved compelling enough that Google announced a similar scheme dubbed Private AI Compute more than a year later.

That might mean more if customers demanded AI device integration, but so far interest in AI hardware has been tepid. AI PCs are not exactly flying off the shelves and Dell reportedly downplayed AI features at CES 2026. 

Part of the issue is that, while local models may eventually be a thing, the trade-offs required for on-device inference, particularly on mobile devices, keep people looking to cloud-based AI services for the time being.

Demand for AI on mobile devices is further complicated by public disinterest in AI phone features. A survey of 2,000 smartphone users in late 2024 found that 73 percent of iPhone customers and 87 percent of Samsung customers said AI added little or no value to their devices.

Samsung, for its part, appears to be pushing back on that narrative by noting that people don't recognize how much they already use AI-powered functions on their phones. A Samsung survey of 2,000 adults conducted by Talker Research found that 90 percent of Americans use AI on mobile devices. But only 38 percent of respondents recognize how many device features have AI elements, such as weather alerts, call screening, autocorrect, voice assistants, auto brightness, and photo editing features.

By the time Apple rewires Siri to work with Gemini, maybe the company will have figured out how AI can add value for customers in a way they recognize and appreciate. ®