Apple unveiled a redesigned Siri AI and a suite of Apple Intelligence tools during its keynote on June 8, 2026 [1].
These updates represent a strategic shift toward on-device artificial intelligence, aiming to make virtual assistants more contextual while maintaining user privacy through private servers.
Speaking from Apple Park in Cupertino, California, CEO John Ternus and SVP Craig Federighi said the tools will integrate with the upcoming iOS 27 [2]. The keynote began at 10 a.m. PT [3].
“The new Siri AI will understand context like never before, making conversations feel natural,” Ternus said [4].
A central part of the announcement was the introduction of Apple Foundation Models. These models are designed to power the new Apple Intelligence tools and provide developers with expanded capabilities on the device [2].
“Apple Foundation Models will be at the heart of the new Apple Intelligence tools,” Federighi said [5].
While some reports suggest the suite may utilize Google Gemini models [6], other sources indicate that Apple Foundation Models remain the core of the system [7]. This hybrid or proprietary approach is intended to give developers unprecedented on-device AI capabilities via iOS 27 [2].
Apple said that the goal of these advancements is to move AI processing away from general cloud services and toward a more secure, on-device environment. This shift allows Siri to handle complex requests by understanding the user's current activity, and history, without compromising data security [8].
““The new Siri AI will understand context like never before, making conversations feel natural.””
Apple's move to integrate foundation models directly into iOS 27 signals a transition from Siri as a simple voice command tool to a proactive AI agent. By prioritizing on-device processing and private servers, Apple is attempting to differentiate itself from competitors who rely heavily on public cloud AI, positioning privacy as a primary product feature in the generative AI race.
