tech
Apple finally ships its AI do-over: Siri AI, a standalone app, and a three-tier privacy stack
Apple unveiled Siri AI at WWDC 2026, a Gemini-powered rebuild with a standalone app, personal context search, and privacy-first cloud architecture.
TL;DR
- Siri AI has been rebuilt from the ground up using Google's Gemini model, integrated as a standalone app and system-wide assistant.
- The new Siri can access personal context from messages, emails, and photos, execute multi-step commands, and search the web.
- A three-tier architecture uses on-device models, Apple's Private Cloud Compute, and Google Cloud for varying request complexities.
- Privacy is a key focus, with assurances that data is used only for requests and not retained or used for model training.
- Apple Intelligence features extend beyond Siri, including enhancements in Photos, Messages, and support for third-party AI models.
- Significant performance improvements are noted across iOS 27, including faster app launches, photo loading, and AirDrop transfers.
- New parental controls include age-appropriate protections, contact approval, communication safety, and customizable Screen Time limits.
- Software releases span iOS, iPadOS, macOS, watchOS, visionOS, and tvOS, with developer betas available now.
- The strategy acknowledges Apple's reliance on Google for a competitive LLM, banking on its vast distribution and privacy architecture.