Google launches Private AI Compute for privacy-centric AI users

Google recently announced a new system called Private AI Compute, which it describes as a way to run its most advanced Gemini models while keeping user data locked up.

Apple already offers a similar feature, introduced in 2024, called Private Cloud Compute.

In a blog post, Google said the feature is designed to balance performance and privacy. Private AI compute would allow AI models to use cloud processing power without giving Google (or anyone else) access to the information being analyzed. According to Google, Private AI Compute operates in a hardware-sealed, verified environment that encrypts all data transfers between the device and its cloud infrastructure.

What do services like private AI compute mean?

Customers who are wary of their personal conversations, personal emails, or sensitive company data being used for AI training often prefer to execute AI processes on the device, which ensures privacy. This way, their data never leaves their computer or phone. However, this is not always possible, as many devices were not built for complex AI processing.

“This approach provides the benefits of a powerful cloud model with the privacy protections of on-device processing,” the company said in its announcement. It is based on Google’s Secure AI Framework and uses its custom TPU along with the new Titanium Intelligence Enclave to enforce end-to-end encryption and security standards, Google said.

Personal AI compute will roll out in select Google products, starting with Magic Queue on the Pixel 10 and updated Recorder app features that use a cloud model to generate smart summaries and suggestions while maintaining conversation privacy.

According to Google, the system aims to support “helpful, personalized, and proactive” AI experiences without requiring users to exchange sensitive information.

As Google’s AI features move beyond simple commands to proactive assistance, Private AI Compute represents what the company calls “the next step in responsible innovation.”