Privacy Computing and Blockchain Integration for Secure Data Flow: Practices and Case Studies by WeBank

With the rapid development of data security regulations, enterprises are increasingly focusing on governance, permission control, and compliance throughout the data lifecycle. This presentation introduces the critical stage of data circulation, covering internal cross‑business sharing, inter‑enterprise data exchange, and cross‑border data usage, highlighting WeBank's explorations and practices.

The agenda includes four parts: (1) new business models under data privacy regulations, (2) blockchain‑enhanced secure multi‑party computation, (3) the construction and practical deployment of a privacy‑computing platform, and (4) a Q&A session.

Data has become a new production factor alongside capital, land, and labor, driving the digital economy. However, data circulation faces compliance challenges because data often contains sensitive information, leading to risks of misuse and leakage. Legal frameworks now restrict plain‑text data export, creating new constraints for data providers, platform users, and data platforms.

Four typical pain points are identified for different roles in data circulation: data providers cannot control downstream usage or obtain fair compensation; platforms face limitations in data handling; and overall, privacy requirements intensify across the data lifecycle.

WeBank outlines three privacy requirements for the data lifecycle: secure storage, trusted transmission, and collaborative production. Traditional technical solutions cannot meet these new security demands.

Four common collaborative production scenarios are described: (a) query‑type where an enterprise queries external parties (e.g., banks querying credit bureaus), (b) intersection‑type for batch queries such as joint marketing or multi‑lending risk detection, (c) prediction‑type involving a data holder and a model holder, and (d) statistical‑analysis‑type for cross‑domain reporting without sharing raw data.

Blockchain is promoted as a trust‑building mechanism for data flow, leveraging immutability, traceability, and multi‑party consensus. Combined with secure multi‑party computation (MPC), it enables encrypted data aggregation and computation, solving the dilemma of plaintext data sharing.

Three technical routes for privacy computing are presented: (1) secure MPC that transforms data into ciphertext for computation and verification, (2) federated learning for AI model training, and (3) trusted execution environments (TEE) as hardware‑based secure enclaves.

The proposed architecture consists of three logical layers: a blockchain‑based trust network storing identities and data metadata, a privacy‑computing network performing ciphertext interactions, and an access network connecting various roles (data owners, model owners, auditors, result receivers, operators) via deployed privacy‑computing nodes.

WeBank's WeDPR solution matrix covers scenarios such as private query, private intersection, joint prediction, joint statistics, anonymous voting, and anonymous ranking. Notable case studies include the 2022 Big Data “Star River” benchmark where WeBank’s privacy‑computing applications were recognized as flagship and excellent cases in medical data collaboration and marine data analysis.

Additional applications include anonymous voting for internal governance and anonymous ranking for auction scenarios, both preserving participant privacy while enabling aggregate results.

The Q&A clarifies that anonymous query returns a set of potential matches without guaranteeing the requester’s data presence, and confirms that privacy‑computing has entered a comprehensive rollout phase across finance, government, public health, digital rights, and other domains.

The presentation concludes with thanks to the audience.