Skip to main content

Android introduces new privacy-friendly sandbox for machine learning data

Android introduces new privacy-friendly sandbox for machine learning data

/

It’s called the Android Private Compute Core

Share this story

Android bot surrounded by padlocks
Illustration by Alex Castro / The Verge

At the I/O developer conference on Tuesday, Google announced a range of new privacy measures, including a new partition within Android to manage machine learning data more securely.

“With Android’s privacte compute core, we’re able to introduce new features,” said Google executive Suzanne Frey in an onstage presentation, “while still keeping your data safe, private, and local to your phone.”

Android’s new Private Compute Core will be a privileged space within the operating system, similar to the partitions used for passwords or sensitive biometric data. But instead of holding credentials, the computing core will hold data for use in machine learning, like the data used for the Smart Reply text message feature or the Now Playing feature for identifying songs. Like the rest of the Android standard, the new system will be open-source and subject to security auditing by third parties.

While neither feature is sensitive in itself, they both draw on sensitive data like personal texts and real-time audio. The partition will make it easier for the operating system to protect that data, while still keeping it available for system-level functions.

“This means that all sensitive audio and language processing happens exclusively on your device and isolated from the network to preserve your privacy,” Google explained in a post announcing the feature.

Despite the name, the Android Private Compute Core is not a separate hardware chip; the partition exists entirely in software. While that lowers the absolute level of data protection, it should also make the system easier to deploy across a range of devices.

The new system was announced alongside a range of other privacy features for Android, including a new privacy dashboard and a new setting for approximate location sharing.