The US government has prohibited the use of differential privacy techniques in Census data collection and reporting. The decision removes a privacy-protection method that adds statistical noise to prevent individual identification.
Differential privacy was introduced in the 2020 Census as a safeguard against re-identification attacks, where individuals could be singled out by cross-referencing public data with other databases. The technique adds carefully calibrated random noise to aggregate statistics while maintaining overall accuracy.
The ban eliminates this protection layer from future Census operations. Differential privacy is widely used in data science to balance privacy and utility—organizations like Apple, Google, and the US Census Bureau have deployed it to prevent privacy breaches without significantly compromising data usefulness.
The move raises concerns among privacy advocates and technologists about Census respondent privacy. Raw Census data remains restricted, but the removal of noise-injection methods reduces protection against sophisticated matching techniques.
The decision affects how demographic, economic, and social data will be released to researchers, policymakers, and the public in subsequent Census cycles.
Security firm Huntress analyzed a real-world intrusion to reveal how threat actors operate once inside a network. The findings show attackers focus on persistence and defense evasion rather than stopping after initial access.
South Korea's Personal Information Protection Commission (PIPC) has fined telecommunications giant KT Corporation KRW 53.979 billion ($39 million) for data protection violations.
JetBrains has disclosed a critical authentication bypass vulnerability in TeamCity On-Premises that allows attackers to execute remote code. The flaw affects the company's continuous integration and deployment platform.