6 ms·
An American Privacy Emergency
Recent and related:
Noise infusion banned from statistical products published by Census Bureau - https://news.ycombinator.com/item?id=48517377 https://news.ycombinator.com/item?id=48517377 - June 2026 (604 comments)
- jmyeet 3mo agoYeah, calling your legislators is going to do precisely nothing [1], just like data centers are almost universally opposed by the communities and the negative externalities are way more real and direct. Yet they keep getting approved anyway. The true crisis here is in the captured political system. In the 1990s in Australia a racist, white supremacist party arose called One Nation through a very weird confluence of events that led a racist fish and chip shop owner by the name of Pauline Hanson to become a member of parliament. It was almost 30 years ago she gave her now famous miaden speech to Parliament [2]. After some scandals, One Nation kind of disappeared for awhile, in part because the conservative coalition (of the Liberals and Nationals) basically adopted the racist platform in the early 2000s where asylum seekers were effectively scapegoated. But weirdly she's back now. Anyway, that part isn't the point. Australia has a preferential voting system, what tends to be called ranked choice voting in the US. You generally have two options on how to vote: you can individually number candidates yourself or you can use the registered preferences for a given party. In this case you put a "1" in Australian Labor Party, Australian Greens or whatever. A lot of people do this so preferences matter. Anyway, One Nation had a strategy of voting gainst the incumbent with preferences. So if it was a Liberal seat, the preference went to Labor and vice versa. This scared the bejsus out of the political establishment such that the opposing political parties gave preferences to each other over One Nation, leading to One Nation getting no seats in Parliament despite getting 10%+ (at its original peak) of the popular vote. My point here is that too many politicians and political parties view their seat as something that belongs to them. In the US primaries are treated largely as a formality by the parties for their anointed candidates. Re-election rates in Congress have sat at 95%+ for decades. What's interesting is that the Demoratic Party is almost in open revolt currently and over the past few weeks, several long-term (10-30 years) incumbents have been primaried by insurgent candidates. Here's a funf act I learned this week. It's been ~18 years since Citizens United basically got rid of campaign spending limits. A third of all the money spent since then has been spent this year on primaries. Thomas Massie has $35M+ spent against him in his primary, making it the most expensive in US history. Many others are in the millions. It's estimated that the total spending for the Senate seat in Maine will push $400M. For one Senate seat. All of this is a long way of saying that the only thing that will work is making these legislators fear they'll lose their cushy positions. And really if somebody has sat in office for 30 years and has nothing really to show for it, it's time for them to go. [1]: https://act.represent.us/sign/problempoll-fba https://act.represent.us/sign/problempoll-fba [2]: https://www.youtube.com/watch?v=p2ypTX9ntTQ https://www.youtube.com/watch?v=p2ypTX9ntTQ
- idle_zealot 3mo agoAgreed: sure, call your representative. If they're cagey or noncommittal, do what you can to get their ass primaried. Every "moderate" will absolutely sell you into a Panopticon.
- onetokeoverthe 3mo agoask tom s. in california if all that spend guarantees anything. more urgent is to repair the broken election process especially in california where it now takes 30+ days to "count" the votes.
- dhotson 3mo agoHeya, fellow Aussie here. Have you ever tried contacting your local MP? I was cynical at first like you, thinking why bother. But when I tried it, turned out I was wrong and I actually had a pretty good experience! The way I see it now, is that MPs aren't always in a good position to get close to the facts, so when you get in touch and tell them what you think.. you're actually giving them a huge gift. It can actually be pretty effective, especially for state/local issues. For federal stuff, sure, might not be as good, but you'll at least get some satisfaction from getting an acknowledgement from their chief of staff or secretary.
- nl 3mo agoAnother Australian here too. Yes, contacting your MP and senators can be very useful, including for federal stuff. It's harder to actually get meetings with Federal members (they spend a lot of time in Canberra) but still worth trying. Also it is very effective to vote for independent senators. You need to pay careful attention to make sure they aren't secretly insane but senators like David Pocock and Jacqui Lambie are very effective (Lambie seems crazy sometimes but she is surprisingly willing to change her mind on issues).
- protocolture 3mo ago>David Pocock Pocock is a blight on the senate. The guy is just a seppo complaint repeating station. I swear he is more in touch with the democrat party than any Australian.
- greyface- 3mo agoPrevious discussion: https://news.ycombinator.com/item?id=48517377 https://news.ycombinator.com/item?id=48517377
- qrush 3mo agoThis post's call to action is talking to your legislators, but it's missing a link to do so. Find yours here: https://www.congress.gov/members/find-your-member https://www.congress.gov/members/find-your-member
- nl 3mo agoRelated: https://news.ycombinator.com/item?id=48517377 https://news.ycombinator.com/item?id=48517377 It's too bad this has become political. I do differential privacy work for GDPR compliance and it's an interesting technology.
- giancarlostoro 3mo ago> I do differential privacy work for GDPR compliance and it's an interesting technology. You mean legislation?
- nl 3mo agoNo? We do GDPR-compliant reporting by using differential privacy to provably remove PII.
- gib444 3mo ago"PII" in the context of GDPR...?
- jolmg 3mo ago>> remove PII
- gib444 3mo agoPII isn't a GDPR term
- disgruntledphd2 3mo agoPersonal data is though, and the two terms have relatively similar meanings.
- gib444 3mo agoUS data privacy terms are US data privacy terms, not GDPR terms. It's sloppy use of language by someone who works in the sector. And businesses wanting to be compliant with GDPR should not be misled with US PII terminology
- aosmith 3mo ago[flagged]
- bogwog 3mo agoWarning: this is AI generated. Don't actually trust it to be secure.
- aosmith 3mo agoHuh? I'm pretty sure I'm a person. But AI would say that too...
- bogwog 3mo agoI was (obviously) referring to the project that you linked to, not saying you are a bot. Github contributors page shows the whole thing was vibe coded: https://github.com/aosmith/cipher/graphs/contributors https://github.com/aosmith/cipher/graphs/contributors
- nickvec 3mo agobogwog is saying that your site/product is obviously vibecoded based off of the frontend
- alexpotato 3mo agoOne of my favorite recently learned facts about Congress: Federally mandated parental leave (paternity and maternity leave) polls at about 80% in favor with the US adult population. This is regardless of political affiliation, by the way. Democrat and Republican voters both support it. Upon reading this, you might be surprised as to why it's NOT federally mandated given how popular it is. One group it's NOT popular with is corporations. And corporations donate a lot of money to politicians. And it's cheaper to donate to politicians who are against parental leave than it is to pay people for that parental leave. I enjoy sharing this b/c it's a reminder that there are groups who spend a lot of time and money to get their way. At first, that might feel overwhelming. You might be surprised to know that when you call your local congressperson, those calls gets tallied b/c they want to know what their constituents care about. So give them a call and let them know.
- hahahaa 3mo agoIf a million people want it and it is worth $100 to them they could create a superpac for this with $100m funding?
- random3 3mo agoYes, apparently there are crowdfunded superpacs exactly for this purpose https://en.wikipedia.org/wiki/Mayday_PAC https://en.wikipedia.org/wiki/Mayday_PAC
- Fomite 3mo agoOne of the things I have found so alarming about a lot of recent revelations is just how cheap congress goes for.
- nelsondev 3mo agoI believe $50-100k
- refulgentis 3mo ago
- throwaway27448 3mo ago[flagged]
- bo1024 3mo agoWhat is the political goal behind this directive? I assume there is some completely non-subtle purpose, but I can't tell what it is.
- kg 3mo agoThe people currently in power want every single byte of data they can get on people who reside in this country so they can divide them up by arbitrary criteria and do with those groups as they please. Depending on the person, that might mean imprisonment, it might mean deportation, it might mean something worse.
- nine_k 3mo agoImprisonment? deportation? these are petty things. The serious danger is polarization and radicalization, turning people against each other, suppressing voices of reason. The ensuing chaos could help pull off another January 6th, only successful. Gerrymandering is one approach, but with more detailed data more interesting things may become possible.
- tjwebbnorfolk 3mo ago> so they can divide them up by arbitrary criteria and do with those groups as they please You mean like by race, gender, and sexual orientation, and using statistical analysis to portray a group of oppressors vs. oppressed, perpetrators and victims? Exactly what was happening via official public policy for the last 4 years while the progressives cheered it on?
- zmgsabst 3mo ago[flagged]
- spiderfarmer 3mo agoYou won't convince anyone with "angry Facebook grandpa" style comments. Please explain to us in great detail how exactly "official public policy" did that, including the laws and changes that were pushed to do so. And explain why, if that were the case, conservatives are doing nothing to stop that. Why would anyone continue to do something bad, just because the people you hate started it?
- onetimeusename 3mo agoThere has been debate among statisticians and political scientists about using differential privacy for census data. 2020 was actually the first Decennial Census that used differential privacy. This is the mandated census done every 10 years that counts population and is used for apportionment. Some have criticized the use of differential privacy.[1][2] But others have argued that coarsening does not protect privacy sufficiently and that differential privacy does not distort apportionment. The political context is unclear. There are lawsuits about whether differential privacy is constitutional. There is also the possibility that citizenship status can be inferred by using multiple census products put together. It's also possible redistricting is at stake although it's unclear to me how getting rid of differential privacy benefits any one party. [1]: https://apnews.com/article/business-census-2020-technology-e701e313e841674be6396321343b7e49 https://apnews.com/article/business-census-2020-technology-e... [2]: https://www.science.org/doi/10.1126/sciadv.abk3283 https://www.science.org/doi/10.1126/sciadv.abk3283
- _alternator_ 3mo agoI'd like to emphasize that coarsening is not just theoretically non-private, a number of attacks that lead to leaking personally identifiable data were demonstrated on the 2010 census. So it's not really a he-said/she-said situation.
- onetimeusename 3mo agoThat's not what I said or implied it is. I said there has been criticism of using differential privacy and linked to it. I also mentioned the reconstruction attacks on coarsening. Those two things can be true at the same time.
- censusnerd 3mo agoThere are glaring factual problems with the florida lawsuit and recent critiques of the 2020 Census DP usage. (See 1.) Whether they're not aware or intentionally malicious, critics who claim the constitutional issue say that DP was applied to state-level counts. But disclosure avoidance (such as DP noise) was never applied to state-level population counts used for congressional apportionment---which is the constitutional purpose. There is no constitutional argument here. Drawing districts using block-level data is a totally different issue (a statutory not constitutional one), and disclosure avoidance was applied to those statistics. 1 https://apnews.com/article/2020-census-bureau-redistricting-statistics-differential-privacy-12a70432d2f55b075abc6e6f8799fe73 https://apnews.com/article/2020-census-bureau-redistricting-...
- wxw 3mo ago> On June 4, 2026, the U.S. Secretary of Commerce issued a directive (DAO 216-26) > DAO-216-26 bans differential privacy and other modern (and not so modern) techniques. It restricts disclosure avoidance techniques to “coarsening,” > DAO-216-26 forbids “noise infusion”, described as “methods that involve modifying a dataset by adding random values, or noise.” > By forbidding noise infusion, the directive bans the disclosure avoidance techniques at the core of dozens of data releases over the last three decades. > Civil servants will do their best to comply with this order while still following the laws that require them to protect the confidentiality of respondents’ data. To balance these competing mandates, they may seek to produce less data or coarsen data so much that it is unusable. Or they might be pushed by political actors to publish data that can be easily unmasked... This current administration is cursed.
- atoav 3mo agoIt is not cursed. It is actively trying to cement its own power like the authoritarian fascists they are. And some of you still pretend if they just cheer enough for dear leader the tide will turn for them as well. As if this all was just a sports game where you have to cheer for your team even if they actively break the rules.
- bdangubic 3mo ago[flagged]
- birdsongs 3mo agoYou're heavily downvoted, and while this admin is absolutely autocratic far beyond anything we've ever seen, and democracy is eroding rapidly: privacy has backslid under every administration. Obama had nearly a decade to revert the patriot act, along with other nefarious shit introduced after 9/11 like PRISM, but did nothing to touch it. America has been creeping towards a surveillance state for decades, regardless of the party in power.
- bdangubic 3mo ago
- zkmon 3mo ago> If followed, this order will destroy the Commerce public data our nation relies on for important decisions, such as where to build necessary services for our community’s well-being So this is not about privacy. Scott sounds like a computer scientist forced (by the American ecosystem) to become a bombastic talker.
- dmvdoug 3mo agoCan’t release data sets if they can’t be sufficiently anonymized so as to avoid Privacy Act/other non-disclosure statutes. Can’t anonymize data sets sufficiently if you’re banned from using the techniques data people use when anonymizing data sets. Not that difficult to follow.
- Sniffnoy 3mo agoThis is guest post by Cynthia Dwork; that's Dwork's writing, not Aaronson's.
- jhhh 3mo agoThe article presents two possible methods to protect privacy in a dataset. It then attacks a theoretical weakness in a contrived scenario in the old method, which intends to incline us to choose the other, newer solution. The article does not describe in detail the newer solution beyond its name. I have have questions that the article doesn't cover in enough detail: 1) Has the coarsening failed in the way described in the article in practice which leaked information? 2) How does the 'other' solution we're expected to desire work? 3) What is an example of the difference in level of detail offered by the newer solution that was not possible when the data needed to be coarsened in practice?
- flowercalled 3mo ago(1) "A Simulated Reconstruction and Reidentification Attack on the 2010 U.S. Census" https://arxiv.org/pdf/2312.11283 https://arxiv.org/pdf/2312.11283 (2) By adding carefully tuned Gaussian noise. In the last 6 years we have also figured out how to add much less Gaussian noise: "The 2020 Census Disclosure Avoidance System TopDown Algorithm" https://arxiv.org/abs/2204.08986 https://arxiv.org/abs/2204.08986 (3) This one is harder to answer, since the Census Bureau aimed to release the same style of statistics as in previous decades. So the goal of 2020 was to release the same statistics with the same error bounds. Evidence suggests they succeeded in doing this. "Evaluating Bias and Noise Induced by the U.S. Census Bureau's Privacy Protection Methods" http://arxiv.org/abs/2306.07521 http://arxiv.org/abs/2306.07521, "Evaluating the Impacts of Swapping on the US Decennial Census" http://arxiv.org/abs/2502.01320 http://arxiv.org/abs/2502.01320
- stymaar 3mo agoCan anyone explain me why the Heritage foundation targeted these statistical techniques? What's the political motive behind it?
- cma 3mo agoProbably related to things like "The Nazis utilized data from routine censuses, tax returns, and municipal police registrations. In Germany, and in occupied countries like the Netherlands, this information was systematically organized. In some instances, IBM technology (via Dehomag punch card machines) was used to tabulate and sort census data to identify individuals of Jewish descent."
- xhkkffbf 3mo agoThe 2020 Census had a number of issues and many of them ended up giving the Democrats more seats in the House than they should have. Oh, and the Census results were due at the White House in December 2016, but somehow they didn't reach the WH until January 21, 2017. Was Differential Privacy specifically involved? I don't think so, but I think it's being lumped into the mix and being blamed for the other issues.
- censusnerd 3mo agoNo disclosure avoidance was applied to the state-level population counts used for congressional apportionment (i.e. how house seats are divvied up). There are plenty of valid critiques of the 2020 Census disclosure avoidance systems, but the Rs who claim that DP was used to manipulate state population counts to give Democrats more seats either have no idea what they're talking about or are being malicious. Also... the first census results typically come out in the same year, e.g. December 2020, not 2016? More detailed demographer data take a bit longer. COVID was really the cause for delayed releases.
- xhkkffbf 3mo agoThe Rs are just confused. The manipulation happened but they didn't need DP to do it. As I said, it wasn't the disclosure avoidance that generated the most discrepancies. It was other forms of adjustments.