Report: Apple Eyes Late February for Gemini-Powered Siri Reveal
Toggle Dark Mode
It’s been two weeks since Apple and Google announced an official partnership to revamp Siri. While rumors that Apple had been exploring such a deal had been circulating for several months, the formal announcement still came as a bit of a surprise.
That wasn’t merely because Apple had chosen to ally itself with one of its biggest tech rivals. After all, considering how much it’s been struggling with Siri, it was inevitable it was going to need to call in some big guns. While Apple reportedly explored several paths, Google and OpenAI are the two 800-pound gorillas in this space right now, and Apple had its reasons for avoiding the latter.
What was more interesting is that Apple was willing to publicly announce this — even somewhat quietly. While many believe the news was more about assuaging investor fears of more Siri delays, it also led to the inevitable concerns that Siri would literally be “powered by Gemini,” effectively becoming a clone of the Android assistant.
Licensing Technology, Not an Assistant
The good news is that’s not the case. Digging below the superficial details of the partnership makes it clear that what Apple is doing is licensing Google’s AI technology, not its out-of-the-box voice assistant. Apple will use the AI large language models (LLMs) that power Gemini — not Gemini itself — to create its own Apple Foundation Models (AFM). It’s dubbing these “version 10” to set them apart from the much weaker “version 1” models that it was working on before enlisting Google’s help.
It’s also telling that Apple has yet to make a public, consumer-facing announcement. What we heard on January 12 was quietly slipped out through business outlets like CNBC and a joint statement published by Google. Apple has yet to put anything in its newsroom or on its website discussing the partnership, so everything we’ve heard beyond the two-paragraph official statement has come from insider information and conjecture.
However, we may not have to wait too long before Apple officially fills in the blanks. By all accounts, the first phase of the new Gemini-powered Siri will debut in iOS 26.4, and the beta cycle for that it likely to begin within the next few weeks. Apple will undoubtedly want to get ahead of the game and announce what it’s doing before the first beta arrives rather than leaving early beta testers to reveal the new capabilities.
In this week’s Power On newsletter, Bloomberg’s Mark Gurman shares some background on Apple’s long road to Gemini, and adds that Apple has some kind of announcement planned for later next month:
Today, Apple appears to be less than a month away from unveiling the results of this partnership. The company has been planning an announcement of the new Siri in the second half of February, when it will give demonstrations of the functionality.
Mark Gurman
Gurman isn’t sure whether that will be a major online event or an invite-only “boutique” briefing like it often holds for members of the press in New York, but in either case, it will ensure that Apple drives the narrative on what the new Siri is — and what it isn’t yet.
That second part may be more important than the first. Ever since it unveiled “Siri 2.0” at its June 2024 Worldwide Developers Conference (WWDC), expectations have run high, and so has disappointment as Apple failed to deliver in iOS 18. If anything, that’s created a divide between those who believe that Apple has something even better in store after all this time and others who have effectively given up on Siri ever being more than it is today.
The Road to iOS 27: From Siri to ‘Campos’
Apple is going to have to manage expectations, as the Siri coming in iOS 26.4 isn’t expected to be as revolutionary as what it has planned for later on. As Gurman revealed last week, Apple is working on a full-fledged chatbot for iOS 27; code-named “Campos,” this will aim at going toe-to-toe with ChatGPT, Gemini, and others. If the reports we’ve heard are accurate, Apple is hoping to unveil this true “next-generation” Siri during its June 2026 WWDC keynote.
Campos could also mark a shift away from Apple’s custom Private Cloud Compute (PCC) to Google servers that use much more powerful tensor chips. That could become a point of tension for Apple’s privacy-focused marketing, but it’s too early to tell where that will land.
Meanwhile, iOS 26.4 will tide us over with a huge jump in intelligence as it moves to the new Apple Foundation Model with 1.2 trillion parameters. That should let Siri analyze on-screen content to take actions based on what it sees, plus gain personal context from apps like Mail, Messages, and Calendar, effectively living up to its 2024 promises for the voice assistant — even if that fulfillment comes a year later than originally planned.
[The information provided in this article has NOT been confirmed by Apple and may be speculation. Provided details may not be factual. Take all rumors, tech or otherwise, with a grain of salt.]
