Apple Taps Google Gemini to Power Siri After Failing to Build Its Own AI

google siri gemini

Apple has finally admitted what industry insiders suspected for months: the company that pioneered the smartphone cannot build a competitive AI on its own. In a joint statement with Google announced Monday, Apple confirmed it will use Google’s Gemini models.

It will also use the cloud technology to power the next generation of Apple Intelligence, including the long-delayed Siri overhaul expected later this year.

This is not a partnership of equals. This is Apple, a company that built its empire on vertical integration and doing everything itself, paying a competitor roughly $1 billion annually for the AI technology it could not develop internally. For a company that prides itself on owning every piece of the technology stack from chip to software, this represents a significant strategic concession.

The Deal Terms

The multi-year partnership gives Apple access to Google’s Gemini models, specifically a 1.2 trillion parameter version custom-built for Apple’s needs. According to Bloomberg reporting, Apple will pay approximately $1 billion per year for this arrangement, though neither company has confirmed the exact figure.

The statement from both companies emphasizes that Apple will independently fine-tune Google’s models, and users will never see Google or Gemini branding. According to The Information, prototypes of the updated Siri show no indication that Google technology powers the assistant. Users will interact with what appears to be pure Apple Intelligence, unaware that their queries route through Google infrastructure.

Apple maintains its existing ChatGPT integration will continue, with a company spokesperson telling CNBC that no changes are planned to the OpenAI agreement. This means Apple will effectively run two external AI partnerships simultaneously, using ChatGPT for complex queries that exceed Siri’s capabilities while relying on Gemini for foundational model work.

Why Apple Made This Choice

Apple evaluated multiple options during what the company called an “extensive evaluation period.” OpenAI was under consideration, as was Anthropic. According to Bloomberg, Apple ultimately rejected Anthropic based on financial terms, finding Anthropic’s fees too high relative to what Google offered.

The decision reflects both Apple’s current capabilities and its near-term ambitions. When speculation about this partnership first emerged, the company was already struggling with Siri development. The “more personalized” Siri promised at WWDC 2024 never materialized on schedule. Apple acknowledged the delays publicly, telling developers the features would take longer than anticipated.

Behind the scenes, Apple has been working on its own trillion-parameter cloud model that could be ready as soon as late 2026. The Gemini partnership functions as a bridge solution, giving Apple time to develop competitive in-house capabilities while delivering the Siri upgrade users have been demanding.

What This Means For Google

For Google, this represents a significant validation moment. The company has been racing to prove its AI capabilities match or exceed OpenAI’s, and Apple choosing Gemini over ChatGPT for foundational model work sends a clear signal to the market. Wedbush analyst Dan Ives called the announcement a “major validation moment for Google.”

The deal also provides Google with extraordinary distribution. Apple devices handle 1.5 billion Siri requests daily across more than 2 billion active devices. Every one of those queries now potentially routes through Google infrastructure, even if users never know it.

Google’s stock rose modestly on the news, but the bigger milestone came when Alphabet’s market capitalization briefly exceeded $4 trillion, making it just the fourth company to reach that threshold after Nvidia, Microsoft, and Apple. Google’s market value now exceeds Apple’s for the first time since 2019.

The OpenAI Problem

While Apple maintains it will continue using ChatGPT for certain features, the Gemini announcement clearly positions Google as the preferred AI partner. OpenAI’s integration with Apple focuses on complex queries beyond Siri’s native capabilities, a narrower use case than powering the entire foundation model layer.

This matters for OpenAI’s growth strategy. The company currently claims over 800 million weekly users, many of whom discovered ChatGPT through word of mouth and direct usage. If Apple users begin associating AI capability primarily with Siri, and Siri becomes meaningfully more capable via Gemini, ChatGPT’s mindshare advantage could erode.

Sam Altman has noted that many people see ChatGPT as synonymous with AI itself. That perception becomes harder to maintain when the world’s most valuable consumer electronics company quietly builds its AI stack on a competitor’s technology.

Privacy And Antitrust Complications

Apple emphasized in its statement that Apple Intelligence will continue running on devices and Private Cloud Compute while maintaining “industry-leading privacy standards.” The company must navigate the tension between its privacy marketing and the reality that Google now processes data for core Apple services.

The timing also coincides with ongoing antitrust scrutiny of the Apple-Google relationship. In August 2024, a federal judge ruled Google acted illegally to maintain its search monopoly, partly through payments to Apple for default search placement. Google paid Apple approximately $38 billion between 2021 and 2022 for that privilege. A December ruling allowed Google to continue making such deals, but the AI partnership adds another dimension to regulators’ concerns about tech industry consolidation.

What Users Will Actually See

The upgraded Siri is expected to roll out later in 2026, likely in spring based on previous reporting. Users should expect more detailed, conversational answers to factual questions, better handling of context and follow-up queries, improved emotional support responses, and stronger performance on relational questions such as identifying family members from context in messages.

Apple is not planning a dedicated chatbot app. The company’s approach treats AI as an enhancement to existing interfaces rather than a standalone product. Siri will become more capable, but Apple will not release a “ChatGPT competitor” as a separate application.

For Apple, the next several months will determine whether outsourcing core AI capabilities to a competitor was prudent pragmatism or strategic surrender. The company has done this before with maps, weather data, and chip manufacturing, eventually replacing external dependencies with internal solutions. Whether it can replicate that playbook in AI, where the technology evolves faster than any previous Apple challenge, remains the most important question facing Tim Cook’s tenure as CEO.