Private AI Compute adds secure server-side memory
Google has outlined an update to Private AI Compute that is intended to provide persistent AI memory across devices while maintaining privacy standards associated with on-device processing. The company says the system will store information in dedicated, encrypted cloud storage, while the cryptographic keys needed to access it remain exclusively on a user’s personal devices.
When a model needs stored information, an authenticated, end-to-end encrypted connection links the device to an isolated cloud environment known as a secure enclave. The enclave temporarily decrypts data in protected memory, processes the request, stores any new context and encrypts the information again. Google says the design uses hardware-enforced enclaves, encrypted channels and per-user databases protected by device-derived keys, leaving the data inaccessible to others, including Google.
Private AI Compute previously handled tasks without retaining context after completion. Alongside an updated technical whitepaper, Google said it will publish a tamper-proof record of the server software so devices can verify its authenticity before transmitting personal data. The company also reported an independent audit by a leading cybersecurity firm and invited privacy specialists to review its architecture and verification methods.
Written by Sakhanda Wire AI from reporting by DeepMind. Facts, quotes and full context are in the original article below.