Apple to let users pick third-party AI models this fall

Apple is preparing to significantly expand its artificial intelligence offerings by allowing iPhone, iPad, and Mac users to choose from multiple third-party AI models when using Apple Intelligence features in the fall releases of iOS 27, iPadOS 27, and macOS 27. According to Bloomberg's Mark Gurman, the new system, internally called "Extensions," will let users access generative AI capabilities from installed apps through features like Siri, Writing Tools, and Image Playground. Apple is already internally testing integrations with Google's Gemini and Anthropic's Claude models, which would join ChatGPT as available options for powering Apple Intelligence system-wide. The Extensions framework will allow AI providers to opt in by adding support through their App Store apps, after which users can set their preferred AI model through the Settings app. Beyond simply choosing a default AI model, users will also be able to assign different Siri voices to different AI providers, meaning responses from Apple's own AI could use one voice while ChatGPT or Gemini-generated responses could use entirely different voices. Apple reportedly aims to position itself as "the platform of choice" for other companies' AI services, though the company is still developing its own language models to remain competitive in the space. The upcoming AI expansion will be formally announced at Apple's WWDC, with the keynote scheduled for June 8. The timing coincides with CEO Tim Cook's impending departure, with incoming executive John Ternus tasked with charting Apple's AI strategy going forward. While Apple is widely perceived as lagging behind competitors in launching new AI services, the company continues to generate significant AI-based revenue and is focusing on integrating AI capabilities into its existing hardware ecosystem rather than building out extensive AI infrastructure.