Toward provably private learning from federated data

Key points

• We announce a new Federated Learning system that provides externally verifiable privacy guarantees while shifting computation to the server to improve training speed, accuracy, and device coverage.1
• Years of research development on anonymization have led to strong differential privacy (DP) guarantees for production models through algorithms like matrix factorization DP-FTRL (MF-DP-FTRL) and distributed differential privacy coupled with Secure Aggregation.1
• In “Toward provably private learning from federated data”, we announce the next generation of our FL system, which leverages Trusted Execution Environments (TEEs) to provide fully verifiable and auditable data anonymization guarantees.1
• The KMS and data processing binaries used in our FL system can be reproducibly built from open source code published in the Confidential Federated Compute Github repository.1

Sources (1)

• [1]Toward provably private learning from federated data
  Google Research Blog · Oct 2, 02:57 PM

  In 2017, Google introduced Federated Learning (FL) a machine learning technique that trains models across decentralized, private data.

  We announce a new Federated Learning system that provides externally verifiable privacy guarantees while shifting computation to the server to improve training speed, accuracy, and device coverage.

Extractive summary: sentences quoted from the sources.