Alibaba’s Secure Data Hub (SDH) leverages MPC, differential privacy and PPML to create a compliant data clean‑room for advertisers, allowing multi‑party data collaboration without exposing raw user data, thereby enhancing privacy, trust and efficiency in the advertising ecosystem.
This article explains data masking techniques—including static (SDM) and dynamic (DDM) methods—detailing how to protect sensitive fields like phone numbers and IDs through replacement, truncation, encryption, randomization, and other schemes, while preserving data utility for testing and analysis.
This article explains why personal data leaks happen, defines data masking (desensitization), compares static and dynamic masking, and details six common masking techniques—invalidating, randomization, replacement, symmetric encryption, averaging, and offsetting—to help developers safeguard privacy.
The article examines how the Personal Information Protection Law creates a new authorization framework and introduces the “computable but not identifiable” concept, arguing that trusted‑computing technologies and controlled environments can reconcile strict privacy safeguards with the data‑driven needs of AI and other industries.
Data masking, also known as data desensitization, transforms sensitive information such as phone numbers and ID numbers using static and dynamic methods—including truncation, randomization, replacement, encryption, and averaging—to prevent privacy breaches while preserving data utility for testing, analysis, and production environments.
