“I realize that many of us have ‘outrage fatigue,’ but this particular outrage hits extremely close to home for the computer science theory community,” wrote Scott Aaronson, a computer scientist at the University of Texas at Austin, in a recent blog post.

Aaronson is referring to a 4 June, 2026 directive from the Department of Commerce that caught agency scientists – and the external research community – by surprise. It bans the Census Bureau and the Bureau of Economic Analysis (BEA), Commerce’s two statistical agencies, from using time-tested methods designed to allow the maximum amount of data to be released from national surveys while protecting the privacy of respondents.

The new policy, Aaronson wrote, “bans essentially all modern techniques for preserving privacy in large datasets, including the 2030 census.” State-of-the-art approaches known as noise infusion – modifying a data set by adding random values – are now off-limits. Instead, the rule says surveys must rely on two older and less effective techniques to avoid disclosing a respondent’s identity.
 

Source https://www.science.org/content/article/scientists-attack-ban-adding-noise-u-s-census-data-and-many-surveys

Another high-profile case has interested scientists. Read now