iOS 27: Siri Integrates Gemini and Claude for Custom AI

⚡

Key Takeaways

1iOS 27 introduces an extension system allowing users to choose the AI that powers Siri, including Gemini and Claude.

2Apple plans to launch this feature in the fall on iOS, iPadOS, and macOS 27, with a presentation at WWDC on June 8.

3Each AI model can have a distinct voice, offering a unique sound personalization for users.

💡Why it matters — This strategic openness from Apple could transform Siri into a flexible AI platform, attracting new tech partners.

Apple Revolutionizes Siri with iOS 27

With the imminent launch of iOS 27, Apple is set to transform Siri into an open platform, allowing users to choose from various artificial intelligence (AI) models to power their voice assistant. Among the available options are Gemini and Claude, which could soon respond to your queries instead of Siri.

An Innovative AI Extensions System

Apple is introducing an AI Extensions system in iOS 27, giving users the ability to select their preferred AI model. This initiative follows the integration of ChatGPT in iOS 18, marking a turning point in Apple's strategy as it acknowledges that Siri can no longer do everything alone. According to Bloomberg, this new system, dubbed "Extensions," will allow users to choose which AI model powers Siri, writing tools, Image Playground, and other Apple Intelligence features.

Unprecedented Openness

In the beta versions of iOS 27, Apple describes the Extensions as a way to access the generative AI capabilities of installed applications on demand, through Apple Intelligence features like Siri. While Google integrates this system into the Gemini app and Anthropic into the Claude app, users will be able to choose which of these models powers their daily interactions.

This announcement fits into a broader strategy. Apple has already struck a deal with Google for Gemini to natively power certain features of Siri and Apple Intelligence. The Extensions go further, giving control to the user rather than Apple to decide which artificial brain operates in the background. ChatGPT, already present since iOS 18 as a fallback solution for queries requiring general knowledge, could also integrate into this new framework.

Launch Expected This Fall

The rollout of this feature is expected this fall, simultaneously on iOS 27, iPadOS 27, and macOS 27. The official presentation will take place during WWDC, which kicks off on June 8.

Unprecedented Sound Personalization

One of the most unexpected new features of this system concerns Siri's sound identity. iOS 27 would allow users to assign a distinct voice to each AI model. Responses generated by Apple's native system would use one voice, while those produced by a third-party model, such as Claude or Gemini, could adopt another. This granularity in personalization illustrates the direction Apple seems to want to take: no longer offering a unique and closed assistant, but building a platform where multiple AI players coexist, each accessible depending on the context or user preferences.

For developers like Google or Anthropic, securing a spot in this system represents privileged access to hundreds of millions of iPhones worldwide. For Apple, it's a way to remain indispensable without having to win the race for models alone. A delicate balance that will be closely scrutinized during WWDC.