Apple is planning to pay Google about $1B a year for an ultrapowerful AI model to help overhaul its Siri voice assistant, according to Bloomberg sources.
Well that one definitely wasn't on my 2025 Bingo Card!
Apple had previously considered using other third-party models like ChatGPT and Claude, but ultimately decided that Google's 1.2 trillion parameter custom Gemini model would best serve its users, at least as an interim solution until Apple's own models are powerful enough, which could be years from now or never. The custom Gemini model is a major advancement over the 150 billion parameter model that Apple currently uses today for its cloud-based version of Apple Intelligence.
Still though, tapping into your biggest competitor for help with AI? That's like Meta adding a Grok chatbot to its mobile apps!
Here's how the arrangement will work:
- Gemini will handle Siri's summarizer and planner functions, which help it synthesize information and decide how to execute complex tasks.
- Some Siri features will continue to use Apple's in-house models. LOL, like “Turn on Flashlight”? Is Apple sure their AI models can handle that one?
- The Gemini model will run on Apple's own Private Cloud Compute servers to ensure that user data remains walled off from Google.
- The partnership won't be visible to Apple users, but will instead be a fully white label solution, unlike the two companies' Safari browser deal, which made Google the default search engine.
Bloomberg notes that the move for Apple “marks an acknowledgement that it has fallen behind on AI — and is now willing to rely on outside technology to catch up.”
As much as I'd like to say something snarky like, “Well iPhone users… might as well switch to Android then!” — I wouldn't bet against Apple in the long run. They've not successfully pioneered any new technology category in decades, but somehow always manage to enter the game late and land on top. Without a doubt, we'll be having a different conversation about Apple's in-house AI efforts within just a few years.