Cupertino, California—Apple has confirmed that its latest Siri upgrade will use Google’s Gemini language models and Nvidia hardware on Google servers. Despite the shift, the tech giant remains steadfast in its promise of user privacy.

For years, Apple has marketed itself as a bastion of privacy, encrypting cloud services so that only users can access their data, even from Apple employees. On-device processing for tasks like image scanning is another way to keep sensitive information local and secure. However, the new Siri AI model requires more computing power than what can be feasibly provided by Apple's own hardware.

Craig Federighi and other executives explained at the WWDC keynote that while user data will not leave their devices, Google servers provide the necessary computational resources for advanced language models. This partnership allows for smarter interactions but raises questions about how far Apple is willing to go to maintain its privacy stance in an increasingly connected world.

The move highlights a broader trend where tech giants are trading off control and data sovereignty for access to cutting-edge technology, all under the guise of user-centricity. Whether this compromise will be seen as acceptable by consumers remains to be seen.