home.social

#pii — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #pii, aggregated by home.social.

  1. Should you leave red herrings about yourself online? · Blog · Alcazar Security

    Short answer: for most people, no. Planting fake jobs, cities, and life details all over the web is a weak default. It rarely wins against systems that ingest public records, commercial data, and whatever you already leaked. It can confuse you on recovery questions, create collateral hassle, and still leave the real trail intact.

    > Interesting read
    #privacy #security #pii

    blog.alcazarsec.com/posts/shou

  2. RE: toot.majorshouse.com/@majorlin

    ““It’s using that… to get information that otherwise would be totally outside of its #jurisdiction … we’re talking about the physical movements of a person who lives in #Canada

    The demand for the man’s location data was included in a request #DHS issued to Google called a customs summons, which is supposed to be used to investigate issues related to importing goods and collecting customs duties.”

    DO NOT SKIP THIS
    Absolute #MustRead
    #Canadians
    #Privacy
    #PersonalData
    #PII
    #InternationalLaw

  3. The #quisling caucus and their Russian and US oligarch backed activists breached the #PII of all Albertan voters. If you're an Albertan voter, #NaheedNenshi wants to know what you want to know about the breach.

    youtube.com/watch?v=3s7XqCdqE6I

    My suggestions: Which of the perpetrators will face criminal prosecution for identity theft (Criminal Code of Canada § 402.2 (1)), trafficking in identity information (§ 402.2 (2)), identity fraud (§ 403), unauthorized use of a computer (§ 342.1), mischief (§ 430), and/or criminal negligence (§ 219)? Which individuals and organizations will face civil liability for violating PIPEDA (federal), PIPA (Alberta, for private entities), and/or POPA (Alberta, for public entities and officials)? Which MLAs will be expelled or recalled? Which appointees will be fired?

    (If complicit Crown attorneys or other officials fail to prosecute or to cooperate in investigations: Alberta has a recall law; organize en masse and use it against every MLA who fails to compel cooperation and investigation or to fire and replace derelict Crown attorneys.)

    What do you want to know?

    #abpoli

  4. 🧠 Bidirectional token-classification — unlike autoregressive LLMs, #PrivacyFilter reads input from both directions simultaneously for deeper context awareness, catching subtle #PII that simple pattern-matching or RegEx rules miss

    ⚡ 1.5B parameter model with only ~50M active parameters (#MoE) — lightweight enough to run on a standard laptop or in a browser, yet achieves ~96–97% F1 score on standard #PII benchmarks #MachineLearning #AI

  5. #OpenAI releases #PrivacyFilter — an open-weight #AI model for detecting & redacting #PII in text. Runs fully locally, no data ever leaves your machine. Apache 2.0 licensed. #opensource

    🧵👇#privacy

    🔍 Detects 8 PII categories in a single forward pass: names, email addresses, phone numbers, physical addresses, URLs, dates, account numbers & secrets (passwords, API keys) — covering virtually all common sensitive data types

  6. OpenAI Privacy Filter: красивая архитектура в суровых условиях русского бенчмарка

    22 апреля 2026 года OpenAI выпустила OpenAI Privacy Filter — открытую модель для поиска и маскирования PII в тексте. На бумаге это выглядит замечательно: небольшая специализированная модель, которую можно запускать локально и без отправки персухи на внешний сервер, длинный контекст и внятная таксономия чувствительных сущностей. Джонов из Айовы или Вошингтон Ди Си она находит замечательно, а что насчет Максима Улугбековича из Нижневартовска? А Галин Палны из Урус-Мартана? После изучения анонса и model card у меня возникло простое человеческое желание: проверить не абстрактный мультиязычный режим, а то, с чем приходится работать в реальной жизни. Я собрал небольшой бенч и хочу поделиться разбором модели и результатами. А они, мягко говоря, в стоке совсем не звездные.

    habr.com/ru/articles/1027266/

    #openai #privacy #pii #персональные_данные #edgedevice #llm #ai #ml

  7. OpenAI Privacy Filter y su Impac…

    El OpenAI Privacy Filter es un modelo avanzado diseñado para detectar y redactar información personal identificable (PII) en textos. Utiliza técnicas de aprendizaje profundo para identificar patrones y contextos en los datos, lo que le permite operar con alta precisión.

    norvik.tech/news/analisis-open

    #Technology #Openai #FiltroDePrivacidad #Pii #Tecnologia #NorvikTech #DesarrolloSoftware #TechInnovation

  8. We finalized our 2025 tax return today.

    My final task was to print out documents from our accountant, physically sign them, scan the signed docs, and upload them to the accountant's secure FTP site.

    No, I don't use any 'convenience' technologies that involve a 3rd party like DocuSign.

    Is there any PII more sensitive than a tax return? Why would anyone allow someone else to have it besides their accountant and the IRS?

    #IRS
    #Taxes
    #PII
    #Privacy

  9. Wisconsinites Can Keep Watching #Porn After Governor Vetoes #AgeVerification Bill

    Evers wrote that the bill doesn’t prevent platforms from giving collected personal data to third parties, such as the government or #dataBrokers. “This is a violation of personal privacy,” he wrote.
    #privacy #pii #security #wisconsin #veto

    404media.co/wisconsin-age-veri

  10. @JohnJBurnsIII
    More and more 'things' seem to be this way.

    "Give us your #PII or you can't <do the thing>."

  11. 𝐏𝐮𝐛𝐥𝐢𝐬𝐡𝐞𝐫𝐬 𝐂𝐥𝐞𝐚𝐫𝐢𝐧𝐠 𝐇𝐨𝐮𝐬𝐞: 𝐑𝐚𝐧𝐬𝐨𝐦𝐰𝐚𝐫𝐞 𝐀𝐭𝐭𝐚𝐜𝐤, 𝐁𝐚𝐧𝐤𝐫𝐮𝐩𝐭𝐜𝐲, 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐂𝐨𝐥𝐥𝐚𝐩𝐬𝐞 𝐨𝐟 𝐓𝐫𝐮𝐬𝐭

    Between late February and early March 2026, #PCH reportedly became the target of a ransomware operation attributed to the group #Anubis, active in the double-extortion ecosystem.

    suspectfile.com/publishers-cle

    #ARB #Data_Breach #FTC #PII #Ransomware

  12. 𝐏𝐮𝐛𝐥𝐢𝐬𝐡𝐞𝐫𝐬 𝐂𝐥𝐞𝐚𝐫𝐢𝐧𝐠 𝐇𝐨𝐮𝐬𝐞: 𝐑𝐚𝐧𝐬𝐨𝐦𝐰𝐚𝐫𝐞 𝐀𝐭𝐭𝐚𝐜𝐤, 𝐁𝐚𝐧𝐤𝐫𝐮𝐩𝐭𝐜𝐲, 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐂𝐨𝐥𝐥𝐚𝐩𝐬𝐞 𝐨𝐟 𝐓𝐫𝐮𝐬𝐭

    Between late February and early March 2026, #PCH reportedly became the target of a ransomware operation attributed to the group #Anubis, active in the double-extortion ecosystem.

    suspectfile.com/publishers-cle

    #ARB #Data_Breach #FTC #PII #Ransomware

  13. 𝐏𝐮𝐛𝐥𝐢𝐬𝐡𝐞𝐫𝐬 𝐂𝐥𝐞𝐚𝐫𝐢𝐧𝐠 𝐇𝐨𝐮𝐬𝐞: 𝐑𝐚𝐧𝐬𝐨𝐦𝐰𝐚𝐫𝐞 𝐀𝐭𝐭𝐚𝐜𝐤, 𝐁𝐚𝐧𝐤𝐫𝐮𝐩𝐭𝐜𝐲, 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐂𝐨𝐥𝐥𝐚𝐩𝐬𝐞 𝐨𝐟 𝐓𝐫𝐮𝐬𝐭

    Between late February and early March 2026, #PCH reportedly became the target of a ransomware operation attributed to the group #Anubis, active in the double-extortion ecosystem.

    suspectfile.com/publishers-cle

    #ARB #Data_Breach #FTC #PII #Ransomware

  14. 152-ФЗ и LLM несовместимы по умолчанию: как мы это исправили без потери качества AI

    Строим AI-ассистента для бизнеса — и обнаруживаем, что каждое сообщение пользователя с персональными данными уходит в Google. Рассказываю, как это исправить, не сломав UX. Когда мы запускали AI-ассистента для квалификации лидов в строительном бизнесе, первый же вопрос от клиента поставил меня в тупик: «А куда уходят персональные данные, которые люди вводят в чат?» Я знал ответ. И он мне не нравился. Пользователь пишет: «Меня зовут Дмитрий, наша компания ООО Ромашка, телефон +7 903 123-45-67, email [email protected] » . Это сообщение в том же виде уходит в Google Gemini API для генерации ответа. Google получает PII — имя, телефон, email конкретного человека. Каждый раз. С каждым пользователем. Для бизнеса в России это три проблемы одновременно. Юридическая. 152-ФЗ требует, чтобы персональные данные российских граждан обрабатывались на территории РФ. Передача данных на серверы Google — даже для обработки, не хранения — это трансграничная передача данных, которая требует уведомления Роскомнадзора и согласия субъекта. Штрафы начинаются от 3 млн рублей. Бизнес-риск. Контактная база клиентов — главный актив отдела продаж. Отдавать её в третьи руки, пусть даже крупной корпорации — вопрос корпоративной гигиены. Этика. Клиент пишет в ваш чат. Он доверяет вам свои данные. Не Google. Задача сформулировалась чётко: большая языковая модель должна вести диалог естественно — обращаться по имени, знать компанию, упоминать email — но никогда не получать реальные персональные данные. Звучит как противоречие. Решение оказалось элегантным.

    habr.com/ru/articles/1015694/

    #информационная_безопасность #персональные_данные #152ФЗ #LLM #большие_языковые_модели #защита_данных #NestJS #Gemini_API #PII #разработка

  15. @alice

    Wrt #PII, It might be a good idea to avoid entering data easily identifiable as trash, and use generators instead. E.g.:

  16. #EURail went off the rails with its data #security incident: eurail.zendesk.com/hc/en-001/s

    Personally identifiable information (name, gender, birth date, passport number, residence...) got stolen and went up for sale on the dark web.

    In a recent email, EU Rail is recommending that clients take extra precaution by updating passwords and don't talk to strangers on the interwebs.

    Why is EU Rail 1) storing 2) unencrypted #pii? Why can't users remove pii from their account after intended use?

  17. Как маскировать персональные данные на изображениях: наш эксперимент с OCR и NER

    Всем привет! Меня зовут Андрей Иванов, я NLP-исследователь в R&D red_mad_robot. Мы разрабатываем систему Guardrails для защиты персональных данных (PII) и фильтрации небезопасного контента. В этой статье расскажу, как мы решали задачу точечного маскирования PII на картинках без обучения специальных визуальных детекторов. Разберём связку оптического распознавания символов (OCR) с NER-моделью, покажем метрики на реальных данных, раскроем ограничения подхода и наши решения для их преодоления.

    habr.com/ru/companies/redmadro

    #ai #llm #ocr #ner #pii #computer_vision #маскирование_данных #обработка_изображений #nlp #rnd

  18. [en] Is #AI "#supercharged #surveillance" #legal? (#USA)

    "... answer is not straightforward."

    "... huge amount of information that the #government can #collect on Americans that is not itself regulated ... by the #Constitution .. #Fourth #Amendment ..."

    "... the government can purchase commercial data ... which can include #sensitive personal information like #mobile #location and web #browsing records."

    "What AI can do is it can take a lot of information, none of which is by itself sensitive, and therefore none of which by itself is #regulated, and it can give the government a lot of powers ...".

    "AI can aggregate ... information to spot patterns, draw inferences ... at massive scale ... law has not caught up with #technological reality".

    technologyreview.com/2026/03/0

    #privacy #pii

  19. OTTAWA - The Privacy Commissioner of Canada today held a press conference regarding the digital attack on Telus Canada's networks and information systems. Telus recently announced that attackers had claimed to have exfiltrated nearly 1 petabyte of company data, including customer data, equivalent to approximately 250,000 DVD movies.

    The Commissioner announced a full investigation will take place. He also indicated that Canadian consumers should not be excessively worried about the breach of their personally identifiable information (PII), as the attackers will still be obligated to follow the requirements of the Personal Information Protection and Electronic Documents Act (PIPEDA), Canada's data privacy law since passage in 2000.

    #Canada #privacy #Telus #hack #hackers #intrusion #exfiltration #PIPEDA #PrivacyCommissioner #security #PII

  20. A Vast Trove of Exposed #SocialSecurity Numbers May Put Millions at Risk of #Identity Theft

    A database left accessible to anyone online contained billions of records, including sensitive personal data that criminals appear to have not yet exploited.
    #privacy #security #pii #identitytheft #ssn

    wired.com/story/a-mega-trove-o

  21. Google's Personal Data Removal Tool Now Covers Government IDs

    #Google on Tuesday expanded its "Results about you" tool to let users request the removal of Search results containing government-issued ID numbers -- including driver's licenses, #passports and #SocialSecurity numbers -- adding to the tool's existing ability to flag results that surface phone numbers, email addresses, and home addresses
    #privacy #security #ssn #identity #pii

    tech.slashdot.org/story/26/02/

  22. NER не про токены: почему span важнее BIO

    NER часто воспринимают как задачу классификации токенов: BIO-теги, последовательности меток, декодирование. Такой взгляд удобен с точки зрения моделей, но плохо отражает то, как NER работает в реальных системах. Сущности - это не токены, а фрагменты текста. Результаты работы NER-систем, как правило, представлены в виде спанов - с явными границами начала и конца (start / end) и типами сущностей. В этой статье мы разберём два уровня разметки в NER: span-level и token-level и покажем, какую роль каждый из них играет в практических пайплайнах.

    habr.com/ru/companies/raft/art

    #ner #named_entity_recognition #аннотация_данных #машинное+обучение #machine_learning #nlp #span #token #персональные_данные #pii

  23. @mikka

    I think the #AI bots are fed and 'know' the entire #API of most #fedi apps, and just walk out each and every pathway of our #SocialGraph, in search for stuff to scrape.

    AI data hunger comes on top of #SurveillanceCapitalism data-is-the-new-oil exploitation of personal information, and our 🥧 #PII is attractively served in this largely wholly unprotected #fediverse of ours.

    Lastly a medical-themed instance more than anything will attract data vultures: 😋 Juicy #privacy-sensitive nuggets.

  24. Large language models are ever more commonly handling sensitive data at scale. 📈

    RAG Servers and MCP Servers serve completely different purposes. The security implications differ just as much, especially around database access. 🔒

    Our latest blog delves into the differences so you can make an informed decision. Check it out 👉 pgedge.com/blog/rag-servers-vs

    #programming #cybersecurity #compliance #pii #hipaa #ccpa #gdpr #privacy #dataprivacy #ai #llm #dataengineering #developers #mcp #rag #postgres

  25. Cybersecurity researchers have disclosed an exposed MongoDB instance containing over 16TB of corporate intelligence and professional data, including PII across billions of records.

    Attribution remains unconfirmed, and while the database was secured after notification, the duration of exposure and potential access are unknown. This incident reinforces how misconfiguration continues to drive large-scale data exposure.

    What technical or governance controls have you found effective in preventing unsecured databases?
    Source: techradar.com/pro/security/16t

    Engage in the discussion and follow TechNadu for objective infosec reporting.

    #InfoSec #DataSecurity #PII #CloudMisconfiguration #CyberRisk #TechNadu

  26. NEW by me:

    From bad to worse: Doctor Alliance hacked again by same threat actor

    databreaches.net/2025/11/18/fr

    This is a bad #databreach in terms of the #PII and #PHI acquired by the hacker, "Kazu," who is about to leak it all.
    Oof.

    Background: I reported on the first breach/attack a few days ago at databreaches.net/2025/11/12/do

    When the CEO claimed it was all secured the same day, the hacker got ticked off and went back in and hacked them again.

    #HealthSec #HIPAA #BusinessAssociate #thirdparty #vendor #hack #ransom #cybersecurity #incidentresponse

    @zackwhittaker @campuscodi @euroinfosec @Hackread