When announced more than a year ago, Apple Intelligence promised a smarter Siri experience, more accurate object recognition, and all without risking user data privacy. But internal turmoil and the loss of researchers to competitors have meant that the promised feature will only be available in early 2026.
In a recent Bloomberg report this morning, Apple is said to have begun testing Google’s Gemini artificial intelligence (AI) model to power several features on its devices. Gemini is believed to power the “World Knowledge Answers” feature, which will provide AI-generated information and summaries for user searches. This will be offered in the form of text, images, videos, and even locations on maps. This feature has been available on Android devices since last year through Gemini Live.
At the same time, Apple will still use its own model for Siri features that require access to user data. They are also still considering the Anthropic Claude and Gemini models to power planning features in Siri.
As a daily Android and Gemini Live user, I find the Google implementation to be very useful – especially Circle to Search – while also being fast. This is a stark contrast to the slow and unfriendly Visual Intelligence in iOS 26.
Meanwhile, the ChatGPT integration in the series over the past year has been a constant cough on the stairs to the point that I have never met an iPhone owner who claims to use it every day.
