How Google’s AI Strategy Led to a $1B Deal with Its Biggest Rival

Google’s long-term AI strategy has paid off in a big way: a $1 billion-a-year deal to put its Gemini AI at the core of Apple’s Siri. This “interim solution” will see Google’s 1.2 trillion parameter model power the “behind-the-scenes” functions of the new “Linwood” assistant.

This partnership, a huge win for Google’s “AI supplier” ambitions, comes after Gemini dominated a “bake-off” against OpenAI and Anthropic. Apple, facing its own AI lag with its 150-billion parameter models, had little choice.

The “Glenwood” project, Apple’s effort to fix Siri, will now rely on this hybrid system. Google’s AI will handle complex “summariser” and “planner” functions, while Apple’s tech will manage simple requests.

Overseen by Craig Federighi and Mike Rockwell, Apple is reluctantly using its rival’s tech to stay competitive. The company is pushing its own teams to build a 1T+ model replacement, but it’s a difficult task.

Privacy is the deal’s cornerstone. The Gemini model will be hosted on Apple’s “walled-off” Private Cloud Compute servers, not Google’s. This ensures Google gets its $1B fee but no access to Apple’s user data.