Apple Debuts New AI Models Built With Gemini
Apple announced new AI models called Apple Intelligence at WWDC 2026, built in partnership with Google using its Gemini technology, emphasizing on-device processing and privacy. The update includes a multi-modal model for speech and image understanding, improved dictation accuracy, and cross-app functionality, though Apple has historically lagged rivals like Google and Anthropic in AI adoption.
Apple unveiled its new AI models, Apple Intelligence, during the WWDC 2026 developers conference on Monday. The models were developed in partnership with Google using its Gemini technology and will integrate into Apple’s ecosystem, particularly enhancing the revamped Siri personal assistant. The company emphasized on-device AI, ensuring smaller models run directly on iPhones, iPads, and Macs without relying on cloud processing. Privacy remains a core focus, with Apple confirming it will not store user data or chat logs with AI. Key improvements include a multi-modal model capable of processing speech and images, along with enhanced dictation accuracy and language understanding. A system orchestrator will coordinate tasks across models, while Apple Intelligence will function seamlessly across various apps. Apple has historically been cautious with AI, avoiding viral moments seen with competitors like Google’s Nano Banana or Anthropic’s Claude Code. However, the company has prioritized thoughtful integration, gradually rolling out AI features into its products. Craig Federighi, Apple’s senior vice president of software engineering, highlighted the user-centric approach during the WWDC keynote, stating that Apple’s AI must be centered around individual needs. The announcement marks a significant step forward for Apple in the AI race, though it remains behind rivals in terms of public visibility.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.