Apple is working hard to fundamentally redesign Siri and make it fit for the future. This involves more than just minor improvements, but a new generation of voice assistant based on modern AI models. To achieve this goal, Apple is currently in discussions with Google about integrating Gemini AI. At the same time, the company is also evaluating its own models and other external partners.
Developments in the field of artificial intelligence are progressing rapidly. Companies like OpenAI and Anthropic have already presented powerful models, and Google is also at the forefront with Gemini AI. Apple doesn't want to be left behind and is therefore completely redesigning Siri. The goal is a more intelligent, personalized, and flexible voice assistant. The discussions with Google demonstrate how seriously Apple is taking this project—even if no final decision has been made yet.
Talks with Google about Gemini AI
According to a Bloomberg report , Apple is in talks with Google about using Gemini AI as the basis for the next version of Siri. Apple has asked Google to develop a specially adapted model to run on its own private cloud compute servers. This approach would combine the integration of powerful AI with Apple's high standards of privacy and security.
Alternative partners: OpenAI and Anthropic
Google isn't the only potential partner. Apple had previously held talks with OpenAI and Anthropic. Both companies have developed models for Apple's servers that are compatible with the planned infrastructure. Anthropic, in particular, was considered a favorite for a time, but the licensing fees demanded led Apple to consider other options. A final decision hasn't been made yet, and no contracts have been signed with either company yet.
Apple's own AI models
In addition to external solutions, Apple is also focusing on developing its own large language models. The company is currently testing several approaches to determine which delivers the best results for users. Two versions of the new Siri are currently in development: one based entirely on Apple's own models, and another that uses external technology.
The second-generation Siri architecture
A key element of the development is the so-called "second-generation Siri architecture." Craig Federighi, Apple's Chief Software Engineer, explained that the voice assistant has been completely rebuilt. With this overhaul, Apple has made precisely the progress needed for modern AI features like personalization, which will be introduced with Apple Intelligence in iOS 18. According to Federighi, the development even goes beyond what was originally announced.
Planned market launch
The schedule for the new Siri version calls for a market launch in spring 2026. Until then, Apple will continue testing various models and decide whether the future voice assistant will be based on its own developments or on technology from Google, OpenAI, or Anthropic.
Apple is testing its own and external AI solutions
Apple has embarked on an ambitious project with the redesign of Siri. Discussions with Google about Gemini AI, previous tests with OpenAI and Anthropic, and the development of its own models show that several paths are open. Which solution ultimately wins out remains to be seen. What is clear, however, is that with the second-generation Siri architecture and the planned AI features in iOS 18, Apple is taking a decisive step toward a significantly more intelligent voice assistant. (Image: Shutterstock / anatoliy_gleb)
