Privacy Computing and Blockchain: Enabling Secure Data Collaboration

In the digital economy, enterprises face challenges of data sharing while protecting privacy and model ownership.

The presentation outlines four parts: background of data privacy, how privacy computing enables data flow, building a privacy computing platform, and practical applications.

Privacy computing, including federated learning, multi‑party computation, and trusted execution environments (TEE), allows data to be processed without exposing raw values, ensuring compliance.

Blockchain provides distributed, traceable, multi‑party consensus, enhancing the trustworthiness of the entire data collaboration workflow.

Use cases such as city flood prediction and bank credit‑card marketing illustrate how encrypted data and models are jointly computed in TEE, with results securely shared.

The proposed platform architecture consists of six layers—from infrastructure to user layer—offering data catalog management, multi‑scenario tasks, multi‑level authorization, and support for various blockchains.

Key advantages include security, performance, ease of use, and flexible extensibility, delivering end‑to‑end trusted data sharing and computation.