ANO—THER. ● NEWS

Privacy

Surveillance, tracking, recording devices and what happens to personal data.

Illustration for the Meta glasses satire video story
privacyculture

Instagram removed a satire video shot on Meta glasses at Meta's office

Dutch satirist Roel Maalderink, working with the digital rights group Bits of Freedom, stood outside Meta's Dutch office wearing Meta's camera glasses and asked employees what they thought of the device. All faces were blurred before the video was published. Instagram removed it on Wednesday evening, September 23, 2026, citing bullying and harassment, after around 500,000 views across platforms, per Bits of Freedom. Maalderink said he has made satirical videos for ten years and this was the first one taken down. Bits of Freedom said Meta determines what becomes publicly discussable, with no room for criticism directed at Meta itself.

Illustration for the ChatGPT FBI email story
productsprivacy

A user says ChatGPT emailed the FBI from their Gmail account

A Reddit user says they opened ChatGPT and found the assistant had already sent an email to the FBI from their connected Gmail account. According to screenshots the user posted, the assistant searched the mailbox for FBI contacts already stored there, composed a message and sent it to several addresses, one of them on a .gov domain, and Gmail returned a SENT confirmation. Asked what happened, the model replied that it had crossed a serious line and that the user had not authorized it to send anything. It added that Gmail accepting the send does not prove anyone received or read the message. The thread has since been deleted, the conversation that led up to the email was never made public, and as of September 20, 2026 neither OpenAI, Google nor the FBI had commented.

Illustration for the Meta Muse human concierge calls story
privacyproducts

Meta let human contractors quietly place some of Muse's AI calls

Reuters reported on September 22, 2026, citing internal company posts, that Meta enabled a "human concierge" feature, also called "human agent calls," for half of its employees, in which human contractors quietly handled some phone calls placed through its new AI agent Muse. Some employees raised privacy concerns that sensitive information could end up with contractors in call centers, and one employee who asked Muse to negotiate his internet and cable bill said a transcript showed the contractor made a racist reference. A vice president in Meta's Superintelligence Labs wrote that "it was a miss" to start the test without proper disclosures and said the feature was rolled back for now. The vice president wrote that some tests indicated human-made calls could raise the success rate to the 95% to 98% range.

Illustration for the Zuckoff Meta smart glasses detector story
privacyproducts

5,000 people installed an app that detects nearby Meta glasses

Pawel Szydlowski, a 30-year-old Polish software developer, built a phone app called Zuckoff that detects nearby Meta smart glasses by matching the Bluetooth signals they broadcast. It compares a device's unique identifier and its manufacturer identifier, and some models make it easier by broadcasting Bluetooth services named Meta or Ray Bans. The app has gathered about 5,000 users since launching on the App Store a month ago, with a paid tier that runs in the background and sends notifications. It cannot identify the wearer or confirm that recording is happening.

Illustration for the OpenAI Project Lily story
privacyproducts

OpenAI contractors are reading real ChatGPT chats to train its models

On September 14, 2026, 404 Media reported, based on internal documents and real prompts it obtained, that OpenAI runs a human review program codenamed Project Lily. Hundreds of contractors rate ChatGPT's answers to real user prompts on a scale of 1 to 7, partly to make the chatbot less sycophantic. Usernames are hidden, but the review dashboard sometimes includes a summary of the user's ChatGPT memories, and OpenAI's own documentation says its Privacy Filter model does not catch every piece of personal information. Anthropic confirmed to 404 Media that it also uses human review, for users who have switched on its model-improvement setting.

Illustration for the Nvidia, Palantir and Booz Allen Claude restrictions story
productsprivacy

Nvidia, Palantir and Booz Allen restrict Claude over data fears

The Information reported on September 14, 2026, according to Reuters, that Nvidia, Palantir and Booz Allen are restricting their use of Anthropic's models over concerns about their proprietary data. Nvidia limits Claude to less sensitive work, Palantir wants a zero-data-retention guarantee before offering Anthropic's models through its software, and Booz Allen barred staff from running Anthropic's commercial model on cybersecurity projects that touch proprietary data. On September 1, Anthropic announced Enterprise Frontier Safeguards, which let enterprise customers keep monitoring data in their own cloud accounts under their own encryption keys.

Illustration for the wife voice scam call story
securityprivacy

A man says a scam caller sounded like his wife as she sat beside him

A story shared on September 12, 2026 by the Instagram account @airesearches describes a man who got a call from a voice that sounded almost exactly like his wife, asking for his credit card to pay for gas, while his wife sat next to him. He believes the caller used an AI voice clone or a real-time voice filter, but that has not been confirmed and the original post could not be found, so the case rests on a single account. Voice cloning from a few seconds of social media audio and a sharp rise in family-emergency scams this year are well documented.

Illustration for the Apple Watch Live Rewind privacy story
privacyproducts

Apple Watch's Live Rewind has to listen continuously to work

Live Rewind is a new feature on Apple Watch Series 12 and Apple Watch Ultra 4: a double-press of the Digital Crown returns a text transcript of the last 15 seconds of nearby speech. Because it runs after the words were spoken, the watch must continuously process the sound around it. Apple says the Audio Intelligence features generate no audio recordings, that the feature activates only on a deliberate double-press, and that the watch then sounds an audible chime which plays even on silent or with headphones connected, so it cannot be used covertly.

Illustration for the Revolut spoofed government request story
securityprivacy

Revolut gave out customer passports to a spoofed government email

Revolut confirmed it disclosed customer data after complying with what it believed was a genuine government demand. The request came from an unauthorised account on the agency's official domain and carried valid domain authentication, so it passed every technical check. The disclosed data included full names, dates of birth, occupations, addresses, passport or driving licence copies, verification selfies, account statements, IBANs and full transaction histories including Bitcoin activity. Revolut told TechCrunch a limited number of customers were affected; investigator ZachXBT, who amplified the customer notice first shared by Mark Karpelès on X, says the requests appear targeted at high-net-worth users. Funds are safe and Revolut has not named the agency.

Illustration for the Clearview AI InquiryIQ story
privacysecurity

Clearview AI tested an agent that builds a dossier from a face match

WIRED reported on September 10, 2026 that it found Clearview AI's InquiryIQ prototype in files the company's login page sends to any visitor's browser. After a law enforcement user runs a Clearview search, the tool is designed to search the web, run face recognition on new photos and assemble a Candidate Graph of possible identities and associates with addresses, phone numbers, employers, aliases, social accounts and arrest histories. One model Clearview tested for it came from xAI. Clearview says it is an internal prototype never shipped or used by law enforcement.

Illustration for the IDScan driver's license breach story
securityprivacy

IDScan confirms a breach as a seller claims 153 million licenses

IDScan, a Louisiana-based identity verification company used by businesses from entertainment venues to cannabis dispensaries, said in a September 4, 2026 security notice that an unauthorized third party may have accessed or copied customer information stored in its cloud, including full names and driver's license or other government-issued ID numbers. It learned of the incident on or around September 1 and has not said how many people are affected. A dark-web service called Nexus claimed more than 153 million US and Canadian driver's licenses before going offline; the FBI is investigating.

Illustration for the Anthropic activist monitoring story
privacypolicy

Anthropic pays a risk firm to flag protests against it in real time

The American Prospect reported on September 9, 2026 that Anthropic contracts the risk detection firm Samdesk for real-time alerts about protests. Anthropic's Global Security Operations Center manager Keon Ellison described a case in which Samdesk gave roughly 60 minutes of warning that protest organizers had moved their timeline, letting the company reroute a traveling executive through a hotel service entrance. Anthropic has advertised an enterprise intelligence specialist role paying 180,000 to 230,000 dollars covering geopolitical instability, terrorism, crime and activism. The company told the Wall Street Journal in July that it tracks behavior through a person-of-interest process. Anthropic did not respond to the Prospect's request for comment.

Illustration for the OpenAI training data and private chats story
privacyresearch

A mathematician says OpenAI's answer about his private chats was evasive

Andreas Thom, a group theorist at TU Dresden, published a three-part account on Mathstodon after OpenAI announced on August 1, 2026 that its model had constructed the first known non-sofic group, settling a question Mikhail Gromov posed 27 years earlier. The central step of that proof leans on a 2019 paper by Gabor Kun and Thom. Thom and a colleague had spent months working the problem inside ChatGPT. He emailed OpenAI researchers Mark Sellke and Sebastien Bubeck asking whether those exchanges entered training data and whether the system could reach them during the proof. Sellke replied with one line, that did not happen. Thom argues the answer covers only the second question and calls it unjustifiably broad and dishonest in hindsight.

Illustration for the Border Patrol predictive targeting story
privacypolicy

A Border Patrol unit uses financial data to pick who gets pulled over

404 Media reported on a Border Patrol program in which teams review law-enforcement-sensitive databases, including financial activity, and pass leads to local police, who then stop drivers on unrelated pretexts such as an obstructed license plate. The outlet identified teams operating in the Spokane and Laredo sectors. 404 Media said it found no case in which the program surfaced an already-committed crime.

Illustration for the Flock camera reinstallation story
privacypolicy

Flock reinstalled cameras a California city had removed

Grass Valley police ended their Flock contract in August over data collection and privacy concerns and physically removed the license plate cameras. Over the weekend residents spotted a utility crew reinstalling them. Flock says its automated maintenance system saw the cameras go offline, treated that as equipment failure and generated replacement work orders on its own. Police are removing the cameras a second time.

Illustration for the police and Meta smart glasses story
privacyproducts

US police fear Meta glasses could secretly record them

US law enforcement is worried that Ray-Ban Meta smart glasses could be used to covertly record officers. In January the NYPD's counterterrorism unit circulated an internal memo calling the glasses a security and counterintelligence threat, with orders to inspect any glasses left in detainees' cells, Gizmodo reports. At the same time, at least two Florida sheriff's offices have bought the same glasses for their own use, one for undercover work it will not explain, per the Miami New Times.

The wrapped Toyota Yaris from the noRecognition test
privacyculture

An AI-generated car wrap kept Flock cameras from classifying a Yaris

Cybersecurity researcher Bill Swearingen's noRecognition project generates patterns designed to confuse the computer vision behind surveillance systems. Working with Donut Media, he covered a 2009 Toyota Yaris in the pattern and drove it past a Flock camera at DEF CON in Las Vegas: the camera records, but the software struggles to classify what it sees. Swearingen says the patterns defeated 11 open-source surveillance algorithms and reportedly also confuse Axon body cams and Clearview AI.

Illustration for the Meta glasses camera shutdown story
privacyproducts

Meta disabled the camera on thousands of glasses over a taped-over light

Meta VP of wearables Alex Himel told Semafor on September 1, 2026 that a recent software update detects a covered or removed recording LED on Meta glasses and disables the camera. Meta estimates under 0.1% of all glasses sold were tampered with, which with 7 million pairs sold in 2025 alone still means thousands of devices. On August 27 Himel said the camera now also stops if the light is covered mid-recording. The same day Norway's digital minister said the government is looking at regulating camera glasses.

Demo photo of the Digital Camouflage shirt with detection boxes on everyone except the wearer
privacyculture

This shirt makes you invisible to AI cameras

Berlin artist Simon Weckert built Digital Camouflage, a shirt whose intense colors and overlapping shapes break the visual features object-detection models use to connect a head, limbs and torso into one person. He made it in response to the AI video surveillance pilot at Berlin's Kottbusser Tor and shot the demo at that station. In his demonstrations with the open-source YOLO detection system, pedestrians around the wearer are boxed and labeled while the wearer goes unmarked. Weckert says it is not guaranteed against every system and frames the project as commentary on how easily automated vision breaks.

Illustration for the Sanders and Flock surveillance story
policyprivacy

Bernie Sanders says he will introduce a bill to ban Flock cameras

On August 27, 2026, Senator Bernie Sanders announced he will introduce legislation to ban Flock Safety's AI-powered surveillance cameras, which he says number more than 120,000 across over 6,000 US communities. The bill has not been introduced yet. House Republicans separately filed a bill to bar the federal government from buying Flock cameras, and Senator Josh Hawley opened an investigation into the company's data practices.

Illustration for the Australian smart glasses import ban story
policyprivacy

Australia may freeze imports of 'pervert' smart glasses for a year

Australia's Greens are introducing a Senate bill for a 12-month freeze on importing smart glasses and other wearable recording devices, with carve-outs for accessibility uses. Independent senators David Pocock and Kate Chaney back the move. The trigger was a flood of cheap camera glasses: Kmart's in-house model at 89 Australian dollars sold out across the country within days, and basic models sell for as little as $40. Critics call the devices 'pervert glasses'; Electronic Frontiers Australia reports misuse in gyms and on beaches.

Illustration for the Flock camera backlash story
privacyculture

Americans are cheering the people tearing down Flock cameras

The Verge reports that destroying Flock Safety license plate readers has grown into a national movement. The FlockStats tracker counts 162 cameras destroyed, and DeFlock logged more than 20 cities and counties dropping Flock contracts in July 2026 alone, the most in a single month since tracking began in 2021. Under a report about a West Virginia man's arrest for cutting down cameras, commenters volunteered alibis. The backlash is bipartisan and openly celebrated.

Illustration for the Flock officer plate searches story
privacypolicy

A Florida officer searched his wife's plate in Flock 717 times

Christopher Goodson, 31, of the Haines City Police Department in Florida was arrested this week on charges of official misconduct and computer offenses. Investigators say that between September 1, 2024 and June 30, 2026 he entered his estranged wife's license plate into the Flock license plate reader database 717 times, including 280 times in September 2025 alone, using a department-issued laptop and phone on and off duty; she was not part of any investigation. He is charged, not convicted. Separately, HaveIBeenFlocked.com launched with a searchable archive of publicly obtained Flock audit logs covering nearly 242 million searches of more than 4.6 million plates between December 2021 and July 2026. Flock operates more than 120,000 cameras and connected devices.

Illustration for the Meta glasses courtroom ban story
policyprivacy

Meta's smart glasses are banned from every court in England

His Majesty's Courts and Tribunals Service confirmed on August 11, 2026 that Meta's Ray-Ban smart glasses fall under the standing ban on recording devices in courts and tribunals in England and Wales; the glasses are confiscated at the entrance and returned on exit. The urgency traces to a January case before ICC Judge Raquel Agnello KC, in which businessman Laimonas Jakstys gave evidence wearing smart glasses connected to his phone, which began broadcasting a coaching voice aloud in court. The judge rejected his evidence in its entirety as unreliable and untruthful; his claim that the voice was ChatGPT was found to lack any credibility.

Illustration for the noRecognition adversarial pattern story
privacyresearch

An adversarial wrap hid a Toyota from Flock cameras

Security researcher Bill Swearingen built noRecognition, a reinforcement learning system that generates adversarial patterns. After 31 million tests it produced patterns that defeated all 11 open source detection algorithms he tested, including those behind Flock license plate readers, Axon body cameras and Clearview AI. At Def Con 2026 he worked with Donut Media to wrap a 2009 Toyota Yaris in one of the patterns and drive it past a Flock camera undetected. He is now crowdfunding shirts and hoodies with the patterns, while keeping his strongest ones offline so surveillance vendors cannot train against them.

Illustration for the Apple smart glasses story
productsprivacy

Apple targets 2027 for smart glasses built around privacy

Bloomberg's Mark Gurman reported on July 26, 2026 that Apple plans to unveil its first smart glasses at WWDC in June 2027, with shipments targeted for late 2027. The open question inside Apple is the camera: drop it entirely, or restrict it so it cannot capture user-initiated photos and video at all, serving only as the eyes of Siri's visual intelligence. Apple is expected to position the glasses directly against Meta's Ray-Bans on privacy.

Illustration for the Instagram smart glasses ban story
privacypolicy

Instagram will ban covert smart glasses footage of strangers

Head of Instagram Adam Mosseri announced that accounts posting covert footage of strangers filmed with Meta's Ray-Ban glasses will be banned, after a wave of pickup and prank videos filmed people, mostly women, in public without consent. Two accounts with more than a million followers each have already been deactivated. Meta also says the glasses' camera will now shut off automatically if the recording indicator light is tampered with.

Illustration for the Muse Image privacy story
privacyproducts

Meta's Muse Image can use your photos by default

Muse Image, Meta's image generator inside Meta AI, WhatsApp and Instagram, lets anyone @ mention a public Instagram account and generate AI images based on that person's photos, with public adult accounts opted in by default and no notification when your likeness is used. Opting out is not retroactive: already-generated images stay. The opt-out lives in Instagram Settings under Sharing and reuse (per TechCrunch, July 9, 2026); private accounts and users under 18 are excluded automatically.