Apple confirms not using Google models, apps or search in Apple Intelligence (asymco.com)

Apple has recently clarified the technology behind its Apple Intelligence platform, specifically its Apple Foundation Models (AFM), emphasizing that it does not utilize any of Google's models, apps, or search infrastructure. During a post-WWDC technical discussion, Craig Federighi, Apple's Senior VP of Software Engineering, assured that Siri and Apple Intelligence operate independently and are not reliant on Google's Gemini models or any of its other proprietary systems. While Apple does leverage Nvidia GPUs hosted in Google’s cloud, he reaffirmed that their models are custom-built for Apple Silicon and trained on Apple’s proprietary data. This announcement is significant for the AI/ML community as it highlights Apple’s commitment to developing its own unique AI infrastructure, reinforcing its stance on data independence and privacy. The Apple Foundation Models consist of five distinct types, including AFM Core for voice capabilities and AFM Cloud for image generation and advanced reasoning tasks. By refining its models with outputs from Google’s frontier models while maintaining an exclusive data approach, Apple is positioning itself as a competitor in the AI field, steering clear of reliance on established tech giants.