标签：#Privacy

The class action lawsuit, filed in Seattle by Virginia resident Charles Sigwalt, claims that Ring's Familiar Faces feature stores images of passersby without consent.

A Google Sheets add-on with 4 million installs has been silently sending your spreadsheet cell data to OpenAI. Hacker News discovered this 9 days ago, when a PromptArmor security report went viral. Last night — when any normal HN story would be decaying into oblivion — it exploded a second time, gaining 59 points and 23.9% in a single day. I track Hacker News every day. I've seen 518 posts come and go over 319 days of systematic monitoring. Most stories follow a predictable death curve: peak on Day 1, bleed points for 2–3 days, then vanish from the Algolia search layer entirely. A post that survives 5 days is exceptional. One that accelerates on Day 9 is something else entirely. Here's the trajectory: 104 → 106 → 148 → 199 → 219 → 247 → (gap) → (gap) → 306 points. Over 9 days, that's a +194.2% total gain. But the real story is the shape of the curve. From Day 5 to Day 6, it added 20 points. From Day 6 to Day 7, roughly 28. Then on Day 9, it jumped 59 points — a single-day increment that's 2–3x the earlier daily gains. 109 comments and counting. This isn't normal HN physics. This is a second wave of attention — the kind that happens when a story percolates through social media and circles back to the search layer with amplified urgency. People didn't just read this and move on. They came back. The vulnerability itself is brutally simple: ChatGPT for Google Sheets, a popular add-on that lets you use GPT inside spreadsheets, sends cell contents to OpenAI as part of every API call. The PromptArmor research documented specific data flows — workbook data that users never intended to share, flowing to OpenAI's servers as part of "context." No breach required. No malicious actor. Just the plugin working as designed, with a data-sharing envelope nobody bothered to read. I've spent 319 days cataloging every AI security signal that hits HN's front page. Patterns emerge when you watch this long. The data is unambiguous: application-layer AI security is the most underserved mark

Thanks to the newly detailed FROST technique, telltale SSD activity can be measured in the browser using simple JavaScript.

Notes on Federated Learning and Differential Privacy 2026-05-31 · privacy-preserving ML Working notes on building federated learning (FL) from scratch, what actually breaks under Non-IID data, and how differential privacy (DP) and secure aggregation fit on top — including the honest negative results that the marketing slides leave out. They follow the implementation in federated-learning-lab (FedAvg / FedProx / SCAFFOLD, DP-SGD, secure aggregation; 33/33 tests, literature cross-validated). 1. What federated learning actually is The data never moves. Instead of pooling everyone's data on one server, each client trains locally and sends model updates to a server that aggregates them. The canonical loop ( FedAvg ) is: Server broadcasts the global model. Each client does a few local SGD epochs on its own data. Each client sends back its updated weights. Server averages the weights (weighted by client data size) → new global model. That's it. The elegance is that raw data stays on-device; the difficulty is that the clients' data distributions are not identical. 2. The Non-IID problem (where FedAvg starts to hurt) FedAvg implicitly assumes every client sees roughly the same distribution. Real clients don't — one hospital sees different cases than another, one phone's keyboard sees different language. Under Non-IID data, each client's local optimum pulls in a different direction, so averaging their updates produces client drift : the global model lands somewhere none of them wanted. Two well-known fixes, both implemented and measured in the lab: FedProx — add a proximal term that penalises drifting too far from the global model. Stabilises training when clients are heterogeneous. SCAFFOLD — track control variates (correction terms) that estimate and subtract the drift direction. More state to communicate, but corrects the bias FedProx only damps. The honest finding worth repeating: on a strongly Non-IID split (e.g. label-skewed MNIST), the fancy methods don't always beat p

Last week a company nobody can name spent $500 million in a single month on Anthropic's Claude API. Not $500K. Not $5M. Half a billion dollars. In one month. Because nobody set a spending limit. Uber burned through its entire 2026 AI coding budget by April . Four months into the year, done. Microsoft quietly cancelled its internal Claude Code licenses and told engineers to go back to GitHub Copilot. All three stories broke within days of each other, and they all point to the same thing. Token-based billing, when given to an ungoverned team, is a financial weapon pointed at your own company. Every prompt, every context window, every agentic loop gets billed. An engineer running Claude Code seriously can rack up $500 to $2,000 a month just by doing their job well. The answer is not stricter policies. The answer is owning the infrastructure and making tokens free. This article breaks down exactly how to do that for a 100-person engineering team for under $1 million, with real 2026 hardware prices and honest tradeoffs. The Root Problem: You Are Renting the Meter When your team uses Claude Code or any external AI API, you do not own anything. You rent compute by the token. The model is not yours. The data leaves your building on every single request. The bill scales with how well your engineers actually use the tool. That last part is the trap. The better your engineers get at using AI, the more it costs you. Uber's Claude Code adoption jumped from 32% to 84% of their 5,000-person engineering org. That is a success story that turned into a budget crisis. Owning the infrastructure flips this completely. The better your engineers get at using AI, the more value you extract from hardware you already paid for. The Solution: Private On-Premise AI The setup is straightforward: Buy GPU server hardware once Download a state-of-the-art open-source model (free) Run an inference server that speaks the OpenAI API format Point Claude Code, Cursor, or any agent at your local endpoint

Plus: A ransomware group is now stealing data in person, BusPatrol wants to hand its license plate surveillance data to the cops, and more.

Border security czar Tom Homan keeps threatening to "flood" New York City with ICE agents. But a new investigation shows that ICE has been quietly ramping up arrests in the New York area already - and disproportionately targeting Latino neighborhoods. The City, a local nonprofit news organization, found 430 street arrests in the metropolitan area […]

As developers, indie hackers, and solo founders, we launch numerous static sites, minimal landing pages, and open-source project documentation blocks. Every single one of these deployments shares a universal prerequisite: a reliable path to gather raw incoming user feedback, inbound sales leads, or bug reports. The traditional path of least resistance has long been to embed a hardcoded HTML <form> inside our page, or worse, expose a standard mailto: link. However, we all know what happens next. Within hours of your app hitting public hosting servers or GitHub, automated asynchronous spam bots find your raw source code, harvest your personal email address, and turn your inbox into a living nightmare. I used to spend hours configuring captchas, writing honey-pot filters, or spinning up custom Serverless Lambda routines just to secure a simple contact form. Eventually, I realized I was fighting the wrong battle. The best way to protect your inbox isn't to build a better shield around your frontend form; it's to remove the form from your code entirely. That is why I built FormCrab.com . 🦀 The Problem: Why Client-Side Forms are a Risk When you embed a custom form or mailto link into your landing page, you are effectively publishing your communication architecture to the world. Spam bots don't even need to render your page anymore; they use basic regex scrapers to crawl through millions of raw static HTML repositories looking for keywords like type="email" or action="..." . Once your endpoint or raw email identity is captured, it is added to bulk programmatic marketing lists. The Trade-Off We All Hate: Option A: Spin Up a Custom Backend. Configuring an Express or Spring Boot API routing layer solely to act as an authenticated SMTP relay. This adds infrastructural complexity and database burdens to what should be a 15-minute frontend project. Option B: Use Form Backends. Even if you use a standard form endpoint handler, you still have to code the frontend UI, handle valida

The US military has long known that cheap fixes could stop location data from exposing its troops. It adopted almost none—and now says adversaries are using the data to target soldiers during a war.

One leading privacy lawmaker said it was time to "start treating the adtech industry as a national security threat."

Analytics is often treated as a reporting feature: a dashboard added after the product already exists. That is usually too late. For software operators, analytics is closer to infrastructure. It is the layer that makes the state of the product visible. Without it, a team cannot evaluate the situation, understand whether the product creates value, or know whether a workflow is improving. That is the reason WebmasterID is built around privacy-first analytics. The goal is not to collect more data than necessary. The goal is to preserve enough signal to make practical decisions without turning measurement into surveillance. Analytics answers operational questions Good analytics starts with plain questions. What happened? Which workflow changed? Which part of the product is used? Where do people leave? Which system events matter? What evidence supports the conclusion? Those questions sound simple, but they are the foundation of product judgment. If the data model cannot answer them, the team is forced to reason from anecdotes, support messages, and internal opinion. Those inputs still matter, but they are not enough on their own. Analytics gives operators a way to compare the current state with the previous state. It makes change visible. It also makes uncertainty visible when the evidence is incomplete. Product value needs evidence A product can look polished and still fail to create value. It can also look unfinished while solving a real operational problem. The difference is usually visible in behavior. Do users return? Do they complete the workflow? Do they avoid a manual step? Does the product reduce confusion? Does it make a business process easier to operate? Privacy-first analytics should help answer those questions without building a profile of every person. In many cases, first-party events, coarse context, workflow state, and careful retention rules are enough. The system does not need to know everything about a user to show whether a product path is working.

I use "what is my IP" sites maybe once a month. Every time I end up on something covered in ads, calling three different tracking APIs, and showing me results I don't fully understand. So I spent a weekend building whatsmy.fyi. The thing I didn't expect: you don't need an IP geolocation API at all if you're on Cloudflare Workers. Every request comes with a cf object that already has your city, country, ISP, TLS version, HTTP protocol, and RTT. Free. Zero latency. The part I enjoyed most was the WebRTC leak test. It checks whether your browser is exposing your real IP through RTCPeerConnection even when you're on a VPN. I ran it on my own setup. It was leaking. Zero logs. Zero storage. Just your data, shown to you. https://whatsmy.fyi

Disclosure: I work on Passlens, a browser-first passport and ID photo maker. This post is about the product decisions behind that workflow, not a neutral review of every tool in the space. A passport photo looks simple until you try to make one that an upload form will actually accept. It is a headshot, yes, but it is also a small chain of constraints: physical size, pixel size, background, head position, print scale, and whatever the destination country's portal decides to reject that week. That is why generic photo editors feel slightly wrong for this job. They can crop. They can resize. They can export. The hard part is not any one of those actions. The hard part is keeping all of them tied to the document rule the user picked. The unit problem For developers, document photos are awkward because two units matter at the same time. A user may need a 2x2 inch passport photo. A visa portal may ask for 600x600 pixels. A print sheet may need 35x45 mm photos at 300 DPI. These are not the same request, but people often treat them as if they are. If the app only thinks in pixels, the print can come out the wrong physical size. If it only thinks in millimetres or inches, the digital upload can be rejected for the wrong pixel dimensions. A good workflow has to keep both ideas alive: the document size and the export target. That is the main reason Passlens keeps presets and print layouts as first-class pieces of the workflow instead of treating them as labels on a crop box. The crop is not the output Another small trap: the crop the user sees is not always the final output. For a digital upload, the crop usually becomes one image file. For printing, the same crop may become several photos arranged on 4x6, A4, or Letter paper with spacing, margins, and optional cut marks. If that print sheet is scaled by the browser or printer dialog, the whole thing is wrong. So the editor needs to separate three things: the face and shoulder crop the finished document-photo size the print s

When regulation becomes theater and encryption becomes window dressing By Vektor Memory — 20 min read It is raining here in the Southern Hemisphere again. It has been raining for three weeks now, nonstop. I’m sitting with my chai coffee, watching out of the window, and thinking about data sovereignty. It is, genuinely, the kind of thing I think about often. The northern hemisphere is winding up for summer. Europe is getting ready for long evenings and beach holidays. I’m quietly jealous. I’ve always wanted to split the year: six months south, six months north. Endless summer. The perpetual warmth of a life lived chasing the sun. But here I am. Chai. Rain. Data. I’ve been turning over one question in particular: why is it that the moment you mention data sovereignty, people immediately reach for GDPR? It’s reflexive, especially among Europeans. Understandable. GDPR is loud, it’s enforced, it has teeth. French, German, and Dutch visitors make up a large disproportionate share of our site traffic at VEKTOR, and the interest in privacy and sovereignty from that audience is intense and genuine. Northern Europeans, by and large, take this seriously in a way that other markets don’t; they are working on ways to disassociate from the cloud around the world. And yet. How many times have we clicked “Accept All” on a cookie banner in the last week? How many times have you scrolled past a privacy policy that runs to forty-two pages? How many times have you handed over your email address, your location, your device fingerprint, your behavioral patterns not because you wanted to, but because there was no meaningful alternative? GDPR created the most sophisticated legal architecture for data rights the world has ever seen. It also created the most sophisticated ritual of consent theater the world has ever performed. That gap, between the law and the lived reality, is what this article is about. Ubiquitous data centre growth image The Reflex Problem When people think of data sovere

The ban for model year 2027 onward began under Biden and has been enacted by Trump.

Not identifying people based on their use of Wi-Fi routers, but identifying people using Wi-Fi signals . This is accomplished through what is known as WiFi sensing , or the use of WiFi signals to infer information about a physical environment. When radio signals like WiFi travel through a space, they interact with the objects and people around them. Those signals can be reflected, scattered, or absorbed. By analyzing how the signal is expected to behave compared with how it is actually received, researchers can infer details about the surrounding environment...

Over 130 companies tangled in sprawling phishing campaign that spoofed a multi-factor authentication system.

Tens of thousands of cameras have failed to patch a critical, 11-month-old CVE, leaving thousands of organizations exposed.

Twitter is blasted for security and privacy lapses by the company’s former head of security who alleges the social media giant’s actions amount to a national security risk.