Damien Desfontaines
desfontain.es · Personal blog · on HN since 2017
Damien Desfontaines, a privacy engineer, writes accessible explanations of differential privacy, anonymization, and the US Census reconstruction attack.
What HN saysHN commenters recommend his census reconstruction explainer as a pretty good account of the concepts, and note his homepage embeds a Nethack game.
25HN people shared it
7linked it in comments
43stories submitted
261story points
4 months agolast shared
2017people per year2026
Most-shared posts
- 1Why differential privacy is awesome (2018)6 people · 7 points · first shared 2018 · discussion on HN
- 2Five things privacy experts know about AI3 people · 5 points · first shared 2025 · 1 comments on HN
- 3K-map, the weird cousin of k-anonymity (2017)2 people · 102 points · first shared 2018 · 5 comments on HN
- 4Demystifying the US Census Bureau's reconstruction attack2 people · 6 points · first shared 2021 · discussion on HN
- 5Differential privacy in practice (easy version)2 people · 4 points · first shared 2018 · discussion on HN
- 6Converting my PhD thesis into HTML2 people · 3 points · first shared 2021 · discussion on HN
- 7Part-time PhD while working at Google: how does it work?2 people · 3 points · first shared 2018 · discussion on HN
- 8Local vs. Global Differential Privacy2 people · 2 points · first shared 2019 · discussion on HN
- 9A friendly, non-technical introduction to differential privacy2 people · 2 points · first shared 2024 · discussion on HN
- 10Almost Differential Privacy2 people · 1 points · first shared 2019 · discussion on HN
- 11Differential privacy in (a bit) more detail2 people · 1 points · first shared 2018 · 1 comments on HN
- 12US bans differential privacy in Census data2 people · 1 points · first shared 2026 · discussion on HN
- 13δ-presence, for when being in the dataset is sensitive1 people · 84 points · first shared 2018 · 5 comments on HN
- 14K-anonymity, the parent of all privacy definitions1 people · 7 points · first shared 2017 · discussion on HN
- 15L-diversity, because reidentification doesn't tell the whole story1 people · 4 points · first shared 2018 · discussion on HN
What HN people wrote when they linked it
p4bl0 · 2020-07-01
My friend Damien's homepage [1] should really go on this list. It embeds a Nethack game on the home page.
[1] Check it out: https://desfontain.es/
TedTed · 2018-10-05
(I wrote the article.)
Excellent point! That's one of the reasons why people understood that syntactic definitions (which try and tell you "how anonymous data is" by looking at the data) were not the right approach, and that you had to look at the mechanism to determine whether the process was sufficiently private.
This line of thinking led to the definition of differential privacy, which completely changed the perspective. I wrote two articles about it, one very simple [1] and the other a bit more detailed [2].
[1] https://desfontain.es/privacy/differential-privacy-awesomeness.html
[2] https://desfontain.es/privacy/differential-privacy-in-more-detail.html
ano-ther · 2026-07-03
From https://news.ycombinator.com/item?id=48517377 / https://desfontain.es/blog/banning-noise.html
> People who were using Census data to actually reconstruct records could no longer do so. Demographers admitted that this was common practice. It's also an open secret that this was done by political operatives as part of gerrymandering efforts.
kccqzy · 2026-06-14
An earlier article on the same blog had some very useful information on how easily the aggregated census data from 2010 (before differential privacy) could be used to reconstruct real data for individuals: https://desfontain.es/blog/us-census-reconstruction-attack.html
I am personally convinced that the reason noise infusion was banned was because powerful people were already reconstructing individual data from census for the purpose of gerrymandering, and they wanted to continue gerrymandering.
drivebycomment · 2021-09-05
https://desfontain.es/privacy/us-census-reconstruction-attack.html is a pretty good explanation on those concepts.
ruuda · 2020-08-13
I found the 6-part series at https://desfontain.es/privacy/differential-privacy-awesomeness.html a very good introduction as well.
majos · 2018-08-13
For those wondering how this compares to differential privacy, the author of this post has also written a short nontechnical post about that: https://desfontain.es/privacy/differential-privacy-awesomeness.html
The theoretical guarantees of differential privacy (e.g. resilience to postprocessing or outside data, which removes the need for most threat modeling) make it the one to beat from a privacy standpoint, IMO.