After Galaxy Phones, Google Gemini to Power Siri on iPhones

Samsung has made Gemini AI chatbot the default assistant on its Galaxy phones, letting users take advantage of Google’s AI capabilities. Now, the company’s arch-rival Apple is also turning to Gemini to improve its own voice assistant, Siri. The Cupertino giant has reportedly been in talks with Google to pay roughly $1 billion annually for a custom version of the Gemini AI model.

Apple picks Google’s Gemini AI for next-gen Siri

For the uninitiated, Gemini is a family of multimodal AI models that can understand and process information from text, images, audio, and video. Apple will mainly use Gemini to handle Siri’s summarizer and planner functions. This will allow the assistant to blend information from multiple sources and decide the best way to execute complex tasks. However, some parts of Siri will still rely on Apple’s in-house AI models.

Vision Pro chief Mike Rockwell and software engineering chief Craig Federighi have led the project. The custom Gemini AI model is quite powerful, with 1.2 trillion parameters. In comparison, Apple’s current AI models use 150 billion parameters. This makes Gemini around eight times more complex, allowing Siri to better understand users’ queries.

Even though Apple plans to use Google’s technology, user privacy remains a top priority. The custom Gemini AI model will run on Apple’s Private Cloud Compute servers. That means personal data will remain on Apple systems, and Google won’t be able to access it. The company has already set up AI servers to run the model.

Earlier this year, Apple tested AI models from other companies as well, including OpenAI and Anthropic. Now, it sees Google’s Gemini as the best option for improving Siri. As for launch, the revamped Siri powered by Gemini may arrive next spring, though the timeline could change. Google’s involvement may remain behind the scenes, meaning users may not even realize that a competitor is powering some of Siri’s intelligence.