MastodonTech.de

SpatialistsDifferential privacy: Being wrong on purpose: How do you protect the <a href="https://mapstodon.space/tags/privacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#privacy</a> of the subjects of <a href="https://mapstodon.space/tags/statistics" class="mention hashtag" rel="nofollow noopener" target="_blank">#statistics</a> and <a href="https://mapstodon.space/tags/data" class="mention hashtag" rel="nofollow noopener" target="_blank">#data</a>? – By adding controlled <a href="https://mapstodon.space/tags/noise" class="mention hashtag" rel="nofollow noopener" target="_blank">#noise</a>. The blog Ironic Sans has an interesting and somewhat funny account of the ramifications of the application of <a href="https://mapstodon.space/tags/differentialprivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#differentialprivacy</a> in the... <a href="https://spatialists.ch/posts/2025/07/01-differential-privacy-being-wrong-on-purpose/" rel="nofollow noopener" translate="no" target="_blank">https://spatialists.ch/posts/2025/07/01-differential-privacy-being-wrong-on-purpose/</a> <a href="https://mapstodon.space/tags/GIS" class="mention hashtag" rel="nofollow noopener" target="_blank">#GIS</a> <a href="https://mapstodon.space/tags/GISchat" class="mention hashtag" rel="nofollow noopener" target="_blank">#GISchat</a> <a href="https://mapstodon.space/tags/geospatial" class="mention hashtag" rel="nofollow noopener" target="_blank">#geospatial</a> <a href="https://mapstodon.space/tags/SwissGIS" class="mention hashtag" rel="nofollow noopener" target="_blank">#SwissGIS</a>

apfeltalk :verified:Apple überarbeitet Siri: „LLM Siri“ soll Neuanfang bringen Apple steht vor einem grundlegenden Umbau seines Sprachassistenten Siri. Ziel ist eine neue, KI-basierte Version namens „LLM Siri“. Diese Entwicklung folgt auf Problem <a href="https://www.apfeltalk.de/magazin/news/apple-ueberarbeitet-siri-llm-siri-soll-neuanfang-bringen/" rel="nofollow noopener" target="_blank">https://www.apfeltalk.de/magazin/news/apple-ueberarbeitet-siri-llm-siri-soll-neuanfang-bringen/</a> <a href="https://creators.social/tags/KI" class="mention hashtag" rel="nofollow noopener" target="_blank">#KI</a> <a href="https://creators.social/tags/News" class="mention hashtag" rel="nofollow noopener" target="_blank">#News</a> <a href="https://creators.social/tags/AppleAIStrategie" class="mention hashtag" rel="nofollow noopener" target="_blank">#AppleAIStrategie</a> <a href="https://creators.social/tags/AppleIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#AppleIntelligence</a> <a href="https://creators.social/tags/AppleSprachassistent" class="mention hashtag" rel="nofollow noopener" target="_blank">#AppleSprachassistent</a> <a href="https://creators.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://creators.social/tags/GenerativeKI" class="mention hashtag" rel="nofollow noopener" target="_blank">#GenerativeKI</a> <a href="https://creators.social/tags/JohnGiannandrea" class="mention hashtag" rel="nofollow noopener" target="_blank">#JohnGiannandrea</a> <a href="https://creators.social/tags/KIApple" class="mention hashtag" rel="nofollow noopener" target="_blank">#KIApple</a> <a href="https://creators.social/tags/LLMSiri" class="mention hashtag" rel="nofollow noopener" target="_blank">#LLMSiri</a> <a href="https://creators.social/tags/PerplexityAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#PerplexityAI</a> <a href="https://creators.social/tags/SiriUpdate" class="mention hashtag" rel="nofollow noopener" target="_blank">#SiriUpdate</a>

Simson Garfinkel@benrothke.bsky.social has written a lovely review of <a href="https://newsie.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> for his "book of the month" column at the RSAC Conference website. He notes that the book is available as an open source download from @mitpress.bsky.social .<a href="https://www.rsaconference.com/library/blog/bens-book-of-the-month-differential-privacy" rel="nofollow noopener" translate="no" target="_blank">https://www.rsaconference.com/library/blog/bens-book-of-the-month-differential-privacy</a>

Marcel SIneM(S)USFehlerbehaftete <a href="https://social.tchncs.de/tags/AppleIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#AppleIntelligence</a>: Abgleich mit <a href="https://social.tchncs.de/tags/iPhone" class="mention hashtag" rel="nofollow noopener" target="_blank">#iPhone</a>-Daten soll helfen | Mac & i <a href="https://www.heise.de/news/Fehlerbehaftete-Apple-Intelligence-Abgleich-mit-iPhone-Daten-soll-helfen-10353093.html" rel="nofollow noopener" translate="no" target="_blank">https://www.heise.de/news/Fehlerbehaftete-Apple-Intelligence-Abgleich-mit-iPhone-Daten-soll-helfen-10353093.html</a> <a href="https://social.tchncs.de/tags/ArtificialIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#ArtificialIntelligence</a> <a href="https://social.tchncs.de/tags/Apple" class="mention hashtag" rel="nofollow noopener" target="_blank">#Apple</a> :apple_inc: <a href="https://social.tchncs.de/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://social.tchncs.de/tags/Datenschutz" class="mention hashtag" rel="nofollow noopener" target="_blank">#Datenschutz</a> <a href="https://social.tchncs.de/tags/privacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#privacy</a>

Vis Lab @ Khoury, NortheasternCongratulations DOCTOR Liudas Panavas on the successful defense of his dissertation "Bridging the Gap: Human Centered Research for Democratizing <a href="https://vis.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> " 🎉 and congrats to advisor <a href="https://vis.social/@codydunne" class="u-url mention" rel="nofollow noopener" target="_blank">@codydunne</a> ❤️ <a href="https://vis.social/tags/HCI" class="mention hashtag" rel="nofollow noopener" target="_blank">#HCI</a> <a href="https://vis.social/tags/DataVisualization" class="mention hashtag" rel="nofollow noopener" target="_blank">#DataVisualization</a> <a href="https://lpanavas.github.io/" rel="nofollow noopener" translate="no" target="_blank">https://lpanavas.github.io/</a>

Ján BogárI just found out about <a href="https://mastodonczech.cz/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> and it's awesome.It's a way of releasing summaries of private data so that almost no info about any individual is leaked. These summaries can be simple, e.g. mean value, or complex, like trained <a href="https://mastodonczech.cz/tags/MachineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#MachineLearning</a> model.E.g. you can train auto-correct on private texts with guarantee that it will not leak during use. How cool is that?<a href="https://mastodonczech.cz/tags/UScensus" class="mention hashtag" rel="nofollow noopener" target="_blank">#UScensus</a> also uses it.Popular summary here: <a href="https://youtu.be/pT19VwBAqKA" rel="nofollow noopener" translate="no" target="_blank">https://youtu.be/pT19VwBAqKA</a>Introductory lecture: <a href="https://youtu.be/9lqd2UINW-E" rel="nofollow noopener" translate="no" target="_blank">https://youtu.be/9lqd2UINW-E</a>

Aurélien BelletThis article does a great job highlighting why DOGE is taking over the federal government so easily: federal systems centralize massive amounts of sensitive data, making them highly vulnerable to insider threats. The article concludes by pointing out that techniques like <a href="https://sigmoid.social/tags/FederatedLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#FederatedLearning</a> and <a href="https://sigmoid.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> could help build more resilient systems 👏 <a href="https://sigmoid.social/tags/Privacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#Privacy</a> <a href="https://sigmoid.social/tags/CyberSecurity" class="mention hashtag" rel="nofollow noopener" target="_blank">#CyberSecurity</a><a href="https://www.nytimes.com/2025/02/21/opinion/musk-doge-personal-data.html?unlocked_article_code=1.yk4.ioNW.2SNQKCzmcCwR" rel="nofollow noopener" translate="no" target="_blank">https://www.nytimes.com/2025/02/21/opinion/musk-doge-personal-data.html?unlocked_article_code=1.yk4.ioNW.2SNQKCzmcCwR</a>

Aurélien BelletTime for a proper <a href="https://sigmoid.social/tags/introduction" class="mention hashtag" rel="nofollow noopener" target="_blank">#introduction</a>! 👋 I'm a <a href="https://sigmoid.social/tags/MachineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#MachineLearning</a> researcher at Inria, France, focusing on <a href="https://sigmoid.social/tags/TrustworthyAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#TrustworthyAI</a>—especially <a href="https://sigmoid.social/tags/decentralization" class="mention hashtag" rel="nofollow noopener" target="_blank">#decentralization</a> & <a href="https://sigmoid.social/tags/privacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#privacy</a>. I design algorithms that learn from <a href="https://sigmoid.social/tags/decentralized" class="mention hashtag" rel="nofollow noopener" target="_blank">#decentralized</a> data (<a href="https://sigmoid.social/tags/FederatedLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#FederatedLearning</a>) without memorizing personal data (<a href="https://sigmoid.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a>), working toward putting <a href="https://sigmoid.social/tags/data" class="mention hashtag" rel="nofollow noopener" target="_blank">#data</a> & <a href="https://sigmoid.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#AI</a> back in people's hands. Maybe one day, these ideas could be used in the <a href="https://sigmoid.social/tags/Fediverse" class="mention hashtag" rel="nofollow noopener" target="_blank">#Fediverse</a>!

Miguel Afonso Caetano"To prevent AI models from memorizing their input, we know exactly one robust method: differential privacy (DP). But crucially, DP requires you to precisely define what you want to protect. For example, to protect individual people, you must know which piece of data comes from which person in your dataset. If you have a dataset with identifiers, that's easy. If you want to use a humongous pile of data crawled from the open Web, that's not just hard: that's fundamentally impossible.In practice, this means that for massive AI models, you can't really protect the massive pile of training data. This probably doesn't matter to you: chances are, you can't afford to train one from scratch anyway. But you may want to use sensitive data to fine-tune them, so they can perform better on some task. There, you may be able to use DP to mitigate the memorization risks on your sensitive data.This still requires you to be OK with the inherent risk of the off-the-shelf LLMs, whose privacy and compliance story boils down to "everyone else is doing it, so it's probably fine?".To avoid this last problem, and get robust protection, and probably get better results… Why not train a reasonably-sized model entirely on data that you fully understand instead?"<a href="https://desfontain.es/blog/privacy-in-ai.html" rel="nofollow noopener" translate="no" target="_blank">https://desfontain.es/blog/privacy-in-ai.html</a><a href="https://tldr.nettime.org/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#AI</a> <a href="https://tldr.nettime.org/tags/GenerativeAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#GenerativeAI</a> <a href="https://tldr.nettime.org/tags/LLMs" class="mention hashtag" rel="nofollow noopener" target="_blank">#LLMs</a> <a href="https://tldr.nettime.org/tags/SLMs" class="mention hashtag" rel="nofollow noopener" target="_blank">#SLMs</a> <a href="https://tldr.nettime.org/tags/Privacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#Privacy</a> <a href="https://tldr.nettime.org/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://tldr.nettime.org/tags/Memorization" class="mention hashtag" rel="nofollow noopener" target="_blank">#Memorization</a>

LavX NewsRevolutionizing Privacy in AI: RAG Meets Differential PrivacyIn a groundbreaking study, Nicolas Grislain introduces a novel approach to enhance privacy in Retrieval-Augmented Generation (RAG) systems. By integrating differentially private token generation, this...<a href="https://news.lavx.hu/article/revolutionizing-privacy-in-ai-rag-meets-differential-privacy" rel="nofollow noopener" translate="no" target="_blank">https://news.lavx.hu/article/revolutionizing-privacy-in-ai-rag-meets-differential-privacy</a><a href="https://mastodon.social/tags/news" class="mention hashtag" rel="nofollow noopener" target="_blank">#news</a> <a href="https://mastodon.social/tags/tech" class="mention hashtag" rel="nofollow noopener" target="_blank">#tech</a> <a href="https://mastodon.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://mastodon.social/tags/RetrievalAugmentedGeneration" class="mention hashtag" rel="nofollow noopener" target="_blank">#RetrievalAugmentedGeneration</a> <a href="https://mastodon.social/tags/EthicalAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#EthicalAI</a>

LavX NewsUnlocking Privacy in Retrieval-Augmented Generation: Meet Sarus DP-RAGIn an era where data privacy is paramount, Sarus DP-RAG emerges as a groundbreaking solution that blends retrieval-augmented generation with differential privacy. This innovative approach not only enh...<a href="https://news.lavx.hu/article/unlocking-privacy-in-retrieval-augmented-generation-meet-sarus-dp-rag" rel="nofollow noopener" translate="no" target="_blank">https://news.lavx.hu/article/unlocking-privacy-in-retrieval-augmented-generation-meet-sarus-dp-rag</a><a href="https://mastodon.social/tags/news" class="mention hashtag" rel="nofollow noopener" target="_blank">#news</a> <a href="https://mastodon.social/tags/tech" class="mention hashtag" rel="nofollow noopener" target="_blank">#tech</a> <a href="https://mastodon.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://mastodon.social/tags/RetrievalAugmentedGeneration" class="mention hashtag" rel="nofollow noopener" target="_blank">#RetrievalAugmentedGeneration</a> <a href="https://mastodon.social/tags/AIPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#AIPrivacy</a>

apfeltalk :verified:Datenschutzbedenken bei neuer Foto-Suche Apple hat in iOS 18, iPadOS 18 und macOS Sequoia neue Funktionen für die Fotos-App eingeführt, die von vielen Nutzer:innen kritisch betrachtet werden. Die „Erweiterte visuelle Suche“, eine KI-gestützte Funktion zu <a href="https://www.apfeltalk.de/magazin/news/datenschutzbedenken-bei-neuer-foto-suche/" rel="nofollow noopener" target="_blank">https://www.apfeltalk.de/magazin/news/datenschutzbedenken-bei-neuer-foto-suche/</a> <a href="https://creators.social/tags/News" class="mention hashtag" rel="nofollow noopener" target="_blank">#News</a> <a href="https://creators.social/tags/Services" class="mention hashtag" rel="nofollow noopener" target="_blank">#Services</a> <a href="https://creators.social/tags/AppleDatenschutz" class="mention hashtag" rel="nofollow noopener" target="_blank">#AppleDatenschutz</a> <a href="https://creators.social/tags/Datenschutz" class="mention hashtag" rel="nofollow noopener" target="_blank">#Datenschutz</a> <a href="https://creators.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://creators.social/tags/ErweiterteVisuelleSuche" class="mention hashtag" rel="nofollow noopener" target="_blank">#ErweiterteVisuelleSuche</a> <a href="https://creators.social/tags/FotosApp" class="mention hashtag" rel="nofollow noopener" target="_blank">#FotosApp</a> <a href="https://creators.social/tags/IOS18" class="mention hashtag" rel="nofollow noopener" target="_blank">#IOS18</a> <a href="https://creators.social/tags/JeffJohnson" class="mention hashtag" rel="nofollow noopener" target="_blank">#JeffJohnson</a> <a href="https://creators.social/tags/KIFunktionen" class="mention hashtag" rel="nofollow noopener" target="_blank">#KIFunktionen</a> <a href="https://creators.social/tags/MacOSSequoia" class="mention hashtag" rel="nofollow noopener" target="_blank">#MacOSSequoia</a> <a href="https://creators.social/tags/Optin" class="mention hashtag" rel="nofollow noopener" target="_blank">#Optin</a>

Simson Garfinkel<a href="https://newsie.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> has a cover. Paperback ISBN: 9780262551656 Pub date: March 25, 2025 Publisher: The MIT Press 244 pp., 5 x 7 in, 22 b&w illus.Advance copies available for reviewers! Email me at simsong@acm.org for info.<a href="https://mitpress.mit.edu/9780262551656/differential-privacy/" rel="nofollow noopener" translate="no" target="_blank">https://mitpress.mit.edu/9780262551656/differential-privacy/</a>

Vis Lab @ Khoury, NortheasternAnd we're off! 🛫 Lots of new faces this year in the Vis Lab @Northeastern @KhouryCollege ! We're excited to continue building community and doing meaningful research together 💙 <a href="https://vis.social/tags/DataVisualization" class="mention hashtag" rel="nofollow noopener" target="_blank">#DataVisualization</a> <a href="https://vis.social/tags/HCI" class="mention hashtag" rel="nofollow noopener" target="_blank">#HCI</a> <a href="https://vis.social/tags/Accessibility" class="mention hashtag" rel="nofollow noopener" target="_blank">#Accessibility</a> <a href="https://vis.social/tags/XAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#XAI</a> <a href="https://vis.social/tags/ML" class="mention hashtag" rel="nofollow noopener" target="_blank">#ML</a> <a href="https://vis.social/tags/DataArt" class="mention hashtag" rel="nofollow noopener" target="_blank">#DataArt</a> <a href="https://vis.social/tags/Health" class="mention hashtag" rel="nofollow noopener" target="_blank">#Health</a> <a href="https://vis.social/tags/Astronomy" class="mention hashtag" rel="nofollow noopener" target="_blank">#Astronomy</a> <a href="https://vis.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a>

Marcel SIneM(S)USInterview zum Schutz von Gesundheitsdaten: "Niemand geht in der Masse unter" | heise online <a href="https://www.heise.de/hintergrund/Interview-zum-Schutz-von-Gesundheitsdaten-Niemand-geht-in-der-Masse-unter-9681689.html" rel="nofollow noopener" translate="no" target="_blank">https://www.heise.de/hintergrund/Interview-zum-Schutz-von-Gesundheitsdaten-Niemand-geht-in-der-Masse-unter-9681689.html</a> <a href="https://social.tchncs.de/tags/Datenschutz" class="mention hashtag" rel="nofollow noopener" target="_blank">#Datenschutz</a> <a href="https://social.tchncs.de/tags/privacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#privacy</a> <a href="https://social.tchncs.de/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://social.tchncs.de/tags/OutputPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#OutputPrivacy</a> <a href="https://social.tchncs.de/tags/Forschung" class="mention hashtag" rel="nofollow noopener" target="_blank">#Forschung</a> <a href="https://social.tchncs.de/tags/research" class="mention hashtag" rel="nofollow noopener" target="_blank">#research</a> <a href="https://social.tchncs.de/tags/AnoMed" class="mention hashtag" rel="nofollow noopener" target="_blank">#AnoMed</a> <a href="https://social.tchncs.de/tags/MaschineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#MaschineLearning</a> <a href="https://social.tchncs.de/tags/ArtificialIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#ArtificialIntelligence</a> <a href="https://social.tchncs.de/tags/DigitalHealth" class="mention hashtag" rel="nofollow noopener" target="_blank">#DigitalHealth</a> <a href="https://social.tchncs.de/tags/Digitalisierung" class="mention hashtag" rel="nofollow noopener" target="_blank">#Digitalisierung</a> <a href="https://social.tchncs.de/tags/digitalization" class="mention hashtag" rel="nofollow noopener" target="_blank">#digitalization</a>

Frederic JacobsInteresting new open-source project from Google to do black-box Differential Privacy testing via divergence optimization over function spaces. Haven't tried it yet but it seems to have promising results: "function-based estimators allow for a better discovery rate of privacy bugs compared to histogram estimation”<a href="https://blog.research.google/2024/02/dp-auditorium-flexible-library-for.html" rel="nofollow noopener" translate="no" target="_blank">https://blog.research.google/2024/02/dp-auditorium-flexible-library-for.html</a> <a href="https://mastodon.social/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> <a href="https://mastodon.social/tags/PETS" class="mention hashtag" rel="nofollow noopener" target="_blank">#PETS</a>

Clément Canonne📢 Next TCS+ talk next week! Wed 11/15, 10:00am PT, Palak Jain (<a href="https://ioc.exchange/@thepalakjain" class="u-url mention" rel="nofollow noopener" target="_blank">@thepalakjain</a>) and Satchit Sivakumar from Boston University will tell us about "The Price of <a href="https://mathstodon.xyz/tags/DifferentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#DifferentialPrivacy</a> under Continual Observation."Details: <a href="https://tcsplus.wordpress.com/2023/11/10/tcs-talk-wednesday-november-15-palak-jain-and-satchit-sivakumar-boston-university/" rel="nofollow noopener" translate="no" target="_blank">https://tcsplus.wordpress.com/2023/11/10/tcs-talk-wednesday-november-15-palak-jain-and-satchit-sivakumar-boston-university/</a>Register (optional): <a href="https://docs.google.com/forms/d/e/1FAIpQLSdvGqPCV2ve1IkUGiARFXTU9o6iWsXwraSJyHAV-mkL7xrArg/viewform" rel="nofollow noopener" translate="no" target="_blank">https://docs.google.com/forms/d/e/1FAIpQLSdvGqPCV2ve1IkUGiARFXTU9o6iWsXwraSJyHAV-mkL7xrArg/viewform</a>

Reuben Binns⁉️It seems like the rationale for <a href="https://someone.elses.computer/tags/differentialprivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#differentialprivacy</a> assumes narrow individual self-interest.Its promise to you is that nothing will be learned from you being part of a dataset that couldn't be learned without you being in it. So even if inferences from the data harm you, this would happen due to others participating anyway.But that rationale only works if you assume people can't imagine co-operating to protect each other by not participating.

Michael VealeThe basic logic of this is extreme horizontal dataset sharding. Imagine a dataset with loads of columns, then imagine each row is held on a different device. Techs such as multi-party computation <a href="https://someone.elses.computer/tags/mpc" class="mention hashtag" rel="nofollow noopener" target="_blank">#mpc</a>, local <a href="https://someone.elses.computer/tags/differentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#differentialPrivacy</a>, can make use of this data. But data is often not visible to the user. Firms claim they do not have to provide rights over it, eg access/portability. Some will put it in the secure enclave of eg a phone; makes it technically very hard to extract (e.g. biometric data).

Michael VealeDid <a href="https://someone.elses.computer/tags/Google" class="mention hashtag" rel="nofollow noopener" target="_blank">#Google</a> ever actually deploy and publicly document what they were analysing with the <a href="https://someone.elses.computer/tags/PROCHLO" class="mention hashtag" rel="nofollow noopener" target="_blank">#PROCHLO</a> protocol in <a href="https://someone.elses.computer/tags/Chromium" class="mention hashtag" rel="nofollow noopener" target="_blank">#Chromium</a>? It's easy to find the technical details of protocols like <a href="https://someone.elses.computer/tags/RAPPOR" class="mention hashtag" rel="nofollow noopener" target="_blank">#RAPPOR</a> (less so for Cobalt in <a href="https://someone.elses.computer/tags/FuchsiaOS" class="mention hashtag" rel="nofollow noopener" target="_blank">#FuchsiaOS</a>), but hard to find what they were actually being used for. Is there a good source on this or all buried in trade secrets and NDAs? (cc <a href="https://octodon.social/@hamed" class="u-url mention" rel="nofollow noopener" target="_blank">@hamed</a> perhaps) <a href="https://someone.elses.computer/tags/PETs" class="mention hashtag" rel="nofollow noopener" target="_blank">#PETs</a> <a href="https://someone.elses.computer/tags/differentialPrivacy" class="mention hashtag" rel="nofollow noopener" target="_blank">#differentialPrivacy</a> <a href="https://someone.elses.computer/tags/browsers" class="mention hashtag" rel="nofollow noopener" target="_blank">#browsers</a>

Frühere Suchanfragen

Suchoptionen

Verwaltet von:

Serverstatistik:

#differentialprivacy