THE CRUNCH
Google DeepMind has introduced a persistent memory layer for its Private AI Compute platform, allowing AI assistants to maintain context across devices while keeping data encrypted and under user control. The system uses hardware-enforced secure enclaves and device-held cryptographic keys to ensure that even Google cannot access the data.
The new architecture resolves a long-standing trade-off between the need for powerful cloud computing and strict on-device privacy. By sealing information in encrypted storage and holding the decryption keys on personal devices, Google aims to provide continuous, seamless assistance without compromising user data.
To build trust, the company is publishing a tamper-proof public record of its server software and sharing the results of an independent cybersecurity audit. This transparency is intended to let users and the broader privacy community verify that the system remains authentic and unaltered before any personal data is sent.