Privacy Computing: How Digital Technologies Address Privacy Protection Pain Points

Recently, the Shanghai Cybersecurity Innovation Research Institute, together with Ernst & Young (China) Consulting, hosted the sixth "Cyber Talk" live event, focusing on the theme “Privacy Technology: How Digital Technologies Address Privacy Protection Pain Points.” The session highlighted the increasing use of privacy computing across smart cities, government data sharing, credit risk assessment, financial anti‑fraud, joint risk control, and precise marketing.

Policy and Commercial Trends – The surge in privacy computing is driven by national policies that treat data as a new production factor and promote data factor marketization. To encourage data sharing while protecting privacy, the State Council has proposed a “raw data stays within domain, data usable but invisible” transaction model, making privacy computing the preferred solution.

Market data shows rapid growth: IDC reported that China’s privacy computing market exceeded CNY 860 million in 2021, with cumulative financing of about CNY 6.5 billion for vendors. Forecasts predict the market will surpass CNY 20 billion by 2025, with a compound annual growth rate of 133.4%.

Technical Analysis – The core privacy computing techniques include:

Secure Multiparty Computation (MPC) : Uses cryptographic protocols to enable parties to compute jointly without revealing raw data.

Trusted Execution Environment (TEE) : Provides a hardware‑based secure enclave where encrypted data is decrypted and processed safely.

Homomorphic Encryption : Allows computation on encrypted data, keeping the data encrypted throughout the process.

Differential Privacy : Adds statistical noise to outputs to prevent inference of individual information.

Federated Learning : Trains models locally on data owners’ devices and aggregates updates, offering a trade‑off between privacy and performance.

These methods aim to achieve the dual goals of “data usable but invisible” and “computable but unidentifiable.”

Challenges and Issues – The panel identified three major challenge categories:

Legal & Regulatory : Unclear data ownership, pricing, and compliance of specific privacy computing techniques under the Personal Information Protection Law.

Technical : No single technique solves all scenarios; trade‑offs exist among accuracy, performance, and security. Standardization, security evaluation, and interoperability among diverse vendor solutions remain unresolved.

Industry Ecosystem : Limited mature applications, scarce valuable data sources, and a fragmented vendor landscape hinder the formation of a robust commercial ecosystem.

The speaker, Wang Lei, General Manager of Ant Group’s Privacy Computing Technology Department, emphasized that Ant Group’s open‑source privacy computing framework “YinYu” will soon be released to foster community participation and accelerate ecosystem development.