Apple announced at WWDC 2026 on 8 June that iOS 27 will introduce a multi-AI Extensions system — a framework that lets users choose their preferred AI assistant from ChatGPT, Gemini, or Claude directly within the operating system. The move makes Anthropic's Claude available on Apple devices for the first time, potentially reaching over 2.2 billion active devices worldwide.
The centrepiece of the AI overhaul is a rebuilt Siri powered by a custom 1.2-trillion-parameter Gemini model licensed from Google for approximately $1 billion per year. Apple took care to emphasise that all Gemini processing runs on Apple Private Cloud Compute, maintaining a strict data separation from Google's own infrastructure — a concession to privacy concerns that would inevitably arise from routing personal assistant queries through a competitor's model.
The AI Extensions system represents a philosophical shift for Apple. Rather than betting exclusively on a single AI provider, Apple is positioning itself as a platform layer that lets users and developers plug in whichever frontier model best suits their needs. For Anthropic, the integration is a significant distribution win — Claude's global market share sits at approximately 8.2 per cent of the AI chatbot market, but with 306 per cent quarterly growth, access to the iPhone's installed base could accelerate that trajectory dramatically.
For context engineers, the WWDC announcements signal that AI model choice is becoming a first-class operating system feature. The era of a single default AI assistant per platform is ending, replaced by a competitive marketplace where users select their preferred model the same way they choose a default browser. Developers building on Apple platforms will need to design for multiple AI backends rather than assuming a single provider.