When Apple began working on a foundational overhaul of Siri, few could have predicted the direction it would take. Internally, Siri had long been seen as underwhelming—a capable voice assistant in 2011, but one that struggled to keep pace with Google Assistant and Alexa as the AI landscape evolved. ChatGPT’s public rise only widened the gap. Faced with its own setbacks in artificial intelligence development, Apple made a decision rarely seen in its playbook: it partnered with a direct rival.
According to a November 2025 Bloomberg report, Apple has struck a deal with Google to use the Gemini AI model to power the next generation of Siri. The company will begin beta testing its iOS 26.4 update, which will introduce this new Apple product and platform strategy change to the public in February 2026. The company, which maintains strict operational control through its vertical integration system, faces both strategic hazards and possible benefits from this decision.
The Backstory: Internal Struggles and Missed Timelines
The decision required extensive consideration before it reached its final outcome. The artificial intelligence research initiative at Apple, which John Giannandrea, who previously worked at Google, leads, experienced several setbacks during its execution. The internal team, which received its assignment to create a new version of Siri using organisational large language models, produced results that did not meet expectations. Bloomberg and Reuters reported that multiple sources identified quality problems together with internal success indicators that had not yet been achieved since Spring 2025.
Apple has maintained its research and development spending at approximately 8 percent of its total revenue. The artificial intelligence spending of Google and Microsoft has grown at an exceptional rate because their cloud systems and data networks support their operations. The AI division of Apple experienced employee attrition during this specific time frame. Ruoming Pang, a prominent researcher, left for Meta. The leadership shuffle culminated in the AI division being moved under Mike Rockwell, previously known for leading Apple’s AR/VR efforts.
Siri Gets a New Brain
The Gemini-powered version of Siri represents more than a technical upgrade. It’s a comprehensive rethinking of what Apple’s voice assistant is capable of. Sources familiar with the development describe features such as conversational memory, on-screen awareness, real-time summarisation, and seamless cross-app interactions. Siri will now be able to generate summaries from emails, suggest responses to messages, or guide users through device settings based on screen context—all powered by Gemini’s advanced capabilities.
The Gemini model from Google was developed to manage various advanced functions, which include reasoning, coding, translation and search enhancement tasks, because it was built with more than one trillion parameters. Apple found the model attractive because its large scale and adaptable features provided a solution to their need for a commercial LLM system after their own development had experienced delays.
Bloomberg reports that Apple will spend about $1 billion each year to access Google’s Gemini system. The amount shows a strategic change that requires substantial financial resources to maintain competitiveness in the artificial intelligence competition. The situation marks an unusual instance where Apple depends on another technology company to support its essential consumer product.
What Will Change for Users
Siri’s new capabilities will initially arrive with iOS 26.4. Users can expect a voice assistant that is faster, more context-aware, and able to handle follow-up questions without losing the thread of conversation. Apple has been demonstrating these features privately, showcasing use cases like summarising a long PDF, generating a reply to a group email, or adjusting settings based on current app activity.
The user experience will be native and branded entirely under Siri, even though Gemini’s technology is at the core. There will be no visible Google branding, similar to how Samsung integrates Gemini under its “Galaxy AI” label. Apple’s emphasis on privacy is preserved through its Private Cloud Compute system, which processes sensitive data on-device and routes more complex AI tasks to secure Apple-run servers.
Importantly, user-identifiable data is not shared with Google under this arrangement. Apple has been careful to underline this point in communications with developers and analysts.
A Transitional Solution
The Gemini integration from multiple reports indicates that it will remain temporary. Apple is developing its own trillion-parameter model while planning to establish its AI infrastructure in-house. The current objective is to implement an assistant with stable performance and strong capabilities during times of market volatility.
Apple will update its existing AI systems, which include the Apple Intelligence initiative and Genmoji and ChatGPT-based features, after assessing which tools remain useful. The analyst Gene Munster indicated that running multiple large language models on different platforms would create expensive and unproductive operations, which leads to the expectation that Apple will merge its AI services into one system.
Implications Beyond Siri
While Siri is the headline feature, the Gemini integration could eventually extend across the operating system. Internal testing has already included Spotlight Search enhancements, Safari query improvements, and app-specific functions. As Apple Vision Pro devices hit the market, natural language interactions powered by AI could become central to the product experience.
This shift also reflects broader market dynamics. Apple’s competitors—including Samsung and Microsoft—have each aligned with major LLM providers to enhance their product offerings. In this case, Apple chose practicality over pride and speed over proprietary exclusivity.
What Remains to Be Seen
There are still open questions. Apple has not confirmed whether the ChatGPT integration launched in 2025 will continue under the new Siri. The terms of the deal with Google remain private, beyond Bloomberg’s financial estimate. It is also unclear whether Apple will eventually make Gemini-powered capabilities available to third-party developers or keep them tightly bound to system-level services.
What we do know is that Apple, for once, is looking outward for answers. And in doing so, it is betting that users care more about capability than the badge behind it.