Apple's Gemini-Powered Siri and Multi-Model AI System Launch at WWDC 2026

Apple's WWDC 2026 keynote, led by CEO Tim Cook in his final appearance, centers on a fully rebuilt Siri now powered by a custom Google Gemini model. The licensing deal is reportedly worth approximately $1 billion per year, giving Apple access to a Gemini variant optimized for on-device and cloud-based processing. The new Siri is designed to handle complex multi-step queries, leverage real-time data, and integrate deeper with third-party apps—all while maintaining Apple's privacy focus through on-device AI processing whenever possible. The model's capabilities include advanced natural language understanding, image analysis, and contextual awareness, positioning Siri to compete directly with Google Assistant and Amazon Alexa.

Beyond Siri, Apple is introducing a new Extensions system that lets users select their preferred AI model for Apple Intelligence features. Instead of being locked into a single provider, users can choose between ChatGPT (OpenAI), Google Gemini, or Anthropic's Claude to power tasks like summarization, image generation, and app automation. The system works through a permissions-based flow: apps request AI access via an API, and the user's chosen model handles the request. This marks a significant strategic shift for Apple, traditionally known for its closed ecosystem, and could set a precedent for multi-model AI integration across consumer devices. Developers will also gain access to these Extensions via an SDK, enabling them to build apps that seamlessly tap into whichever AI engine the user prefers.