Published on [Permalink]
Reading time: 5 minutes
Posted in:

Apple didn't get gobbled by Google

Many are saying it. Google Gemini is inside Apple’s iPhone AI. Apple has been gobbled by Google. The companies' phones are barely distinguishable now.

None of that is right.

The confusion is understandable. Google uses the word “Gemini” to mean at least two different things: their consumer AI assistant app and their family of underlying AI models. When people hear that “Gemini is powering Siri AI,” they picture the Google Gemini app somehow inside an iPhone. That is not what is happening.

Here is what Apple actually announced. From their own newsroom:

“These new capabilities are powered by the next generation of Apple Foundation Models, custom-built in collaboration with Google and its Gemini models for deeply integrated Apple Intelligence experiences.”

Read that carefully. Apple Foundation Models. Custom-built. In collaboration. Google’s technology is one ingredient in an Apple-engineered stack, not a product drop-in.

Your data is not going to Google

Apple’s privacy architecture is clear on this. When Private Cloud Compute handles your requests, your personal data is not stored and is not accessible to Apple or anyone else - Google included. Apple’s newsroom is explicit: users' private data, messages, contacts and screen context stay within Apple’s privacy envelope.

If you use the Google Gemini app on your iPhone, your data goes to Google. If you use Siri AI, it does not. These are categorically different things.

The experience is different too. Ask Google Gemini a question and ask Siri AI the same question. You will get the same core facts on factual matters, but the response, the personality, the system-wide integration, the ability to act on your personal context - that is all Apple. Apple controls the interface, the intelligence layer and what the AI can and cannot see.

We have been here before

Ars Technica’s 20-year retrospective on the Intel Mac landed this month, and the timing is instructive.

For fifteen years, every Mac ran Intel processors - the same Intel that powered every Windows PC on the planet. Intel was inside both. Nobody ever confused a MacBook with a Dell. Apple controlled the experience. Intel did the grunt work under Apple’s direction. Apple never put “Intel Inside” stickers on its computers, because the processor was a detail. What mattered was everything Apple built around it.

The same logic applies here. Google is co-engineering part of Apple’s AI foundation. Apple builds everything that matters around it.

Note what Apple’s then-CEO Tim Cook said when announcing the Apple Silicon transition in 2020: “When we make bold changes, it’s for one simple yet powerful reason. So we can make much better products.” He could have said the same thing about the original Intel switch in 2005. The language is identical because the play is identical.

Apple has done this five times

This is not a one-off compromise. It is Apple’s standard operating procedure.

The Apple II ran a MOS 6502. The original Mac ran Motorola processors. The PowerPC era was a collaboration between Apple, IBM and Motorola. Then Intel. Then Apple Silicon, built on technology Apple acquired when it bought P.A. Semi in 2008 and spent more than a decade scaling up.

Each transition followed the same arc. Apple leaned on the best available external chip technology, built its experience on top, and eventually internalised the capability when the economics and engineering made it possible. There is no reason to think the AI era will be different.

Two years ago, Apple extended Siri with a gateway to ChatGPT - a relatively thin integration. This year, they are more deeply co-engineering the foundation with Google. The direction of travel is toward more Apple-made capability over time, not less. That is the historical pattern and there is nothing in the current situation to contradict it.

The experience is the product

There is a sharper point buried in all of this. The AI industry is at an inflection where the underlying models are commodifying faster than anyone expected. OpenAI, Google, Anthropic, Meta - and now open-weight models from DeepSeek in China and Mistral in France - the foundation model capabilities are converging globally. The differentiation is increasingly in what surrounds the model: the integration, the context, the interface, the trust relationship with the user.

Apple is better positioned than almost anyone on exactly those dimensions. They have the device, the operating system, the app ecosystem and fifteen years of privacy credibility. Google’s model technology is one input into that. It is a significant input, and Apple could not have built the best version of Siri AI without it - the failed 2024 demos made that plain enough. But the model is not the product. The experience is the product.

Google provides the engine. Apple builds the car.

When Apple ran Intel chips, nobody called a MacBook a “PC with a nice case.” The processor was a means to an end. The same is true here. There is Google technology inside Siri AI - but the interface, the privacy model, the system-wide integration and the overall experience are Apple’s. That is not being gobbled. That is Apple doing what Apple has always done: finding the best available technology, building its experience on top, and keeping the relationship with the user entirely to itself.


Sources:

✍️ Reply by email