Apple launches Siri AI standalone app with three-tier privacy

At its annual Worldwide Developers Conference on Monday, Apple introduced Siri AI, a standalone application that represents the most dramatic transformation of its voice assistant in 15 years. The system has been completely rebuilt using a custom Google Gemini model, marking a major departure from previous iterations. The WWDC 2026 keynote at Apple Park also carried special significance as it was Tim Cook’s final appearance as CEO, with the executive set to pass leadership to his successor.

The new Siri AI app arrives with a three-tier privacy framework, designed to give users granular control over how their data is handled. This structure allows the assistant to process requests entirely on-device, through a secure cloud relay, or via direct connection to Apple’s servers, depending on the sensitivity of the task. Apple emphasized that the privacy architecture is foundational to the assistant’s redesign, not an afterthought.

By partnering with Google on the Gemini model, Apple has effectively acknowledged that third-party AI expertise was necessary to achieve the level of performance and contextual understanding users now expect. The integration is said to enable more natural conversations, better task completion, and deeper integration with third-party apps.

The launch of Siri AI as a standalone app, rather than a feature buried within the operating system, signals Apple’s intent to position the assistant as a central, proactive platform rather than a passive tool. The move comes as competitors have raced ahead with more sophisticated voice agents, and Apple appears determined to reclaim its leadership in the category.

With Tim Cook’s tenure ending, the introduction of Siri AI may be remembered as one of the most consequential product decisions of his final year, setting the stage for the company’s next chapter under new leadership.