For months, Apple has been painted as the slow mover in the AI race—carefully plotting while competitors sprint ahead. But in a twist that feels straight out of the classic tortoise-and-hare story, Apple may have found a powerful shortcut: teaming up with Google’s Gemini AI to dramatically upgrade Siri and strengthen Apple’s overall AI experience, even as some critics fume about what it means for the industry.
At the center of the news is Apple’s decision to use Google Gemini to power the next generation of its on-device Foundation Models. The partnership is expected to play a key role in bringing a significantly improved Siri to users, potentially arriving with a later iOS 26.4 update. That revamp is aimed at delivering features many users have been waiting on for a long time, including in-app actions, stronger awareness of personal context, and the ability to understand what’s on your screen.
In practical terms, that means Siri should become far more useful in real day-to-day situations—less like a voice tool that answers simple questions, and more like an assistant that can actually do things across apps while understanding what you’re trying to accomplish.
One of the biggest details is how Apple reportedly plans to run Gemini at scale. The company is said to be deploying a massive custom Gemini model—around 1.2 trillion parameters—on Apple’s own cloud servers. This supports Apple’s broader approach to “Private Apple Intelligence,” where smaller tasks can be handled directly on the device, while more complex requests are securely offloaded to Apple’s private cloud. The key point in Apple’s pitch is privacy: data is processed using encrypted and stateless methods before AI inference, aiming to limit how much personal information is exposed during processing.
Unsurprisingly, the deal has drawn criticism, including from Elon Musk, who called the arrangement an “unreasonable concentration of power” for Google given its existing dominance in areas like Android and Chrome. Musk is also not a neutral observer here, having already taken legal action against other major AI partnerships that he argues stifle competition.
Still, the bigger business reality is that Google’s upside from this arrangement appears relatively modest. Reports suggest Google is earning about $1 billion per year in licensing fees—significant, yes, but not a transformational win for a company of Google’s scale.
Apple, however, stands to gain something far more valuable than a simple feature upgrade: a stronger network effect around Siri as the face of its AI future.
Here’s why. If you ask Siri to book a restaurant reservation, summarize your messages, or carry out actions inside an app, you’re not going to tell your friends that “Gemini did it.” You’ll say Siri did it. From the user’s perspective, the intelligence is Apple’s—because Apple controls the interface, the workflow, and the experience. The underlying AI model becomes invisible infrastructure.
That dynamic is Apple’s ace card. Gemini can serve as a powerful bridge, letting Apple deliver a smarter, more capable Siri right away, while Apple continues working behind the scenes to raise its own AI capabilities to the level it ultimately wants. In the meantime, every successful Siri interaction boosts trust in Apple’s ecosystem, increases engagement, and makes the iPhone’s AI experience feel more essential—exactly the kind of advantage Apple has always been best at turning into long-term momentum.
