Cybersecurity

Samsung Smart TV Data Leak: 150,000 Viewing Records Exposed in n0n Breach

Published By Shivam Sharma8 min read
Smart TV glowing in a dark living room, representing the Samsung viewing-data leak

What n0n dumped

A ransomware group called n0n dumped about 220MB of data it claims belongs to Samsung Electronics on its dark web leak site, after an extortion countdown expired, according to Cybernews, whose researchers investigated the exposed files. The group had listed Samsung Electronics Co., Ltd. as a victim on its leak site on 14 November 2026, according to threat-monitoring services cited by reef-sentinel.net.

n0n is a relatively new ransomware operation that first attracted attention in mid-September 2026, according to Cybernews. Cybersecurity firm SOCRadar found that n0n’s Python file server was left exposed on the open internet, and that the attackers heavily rely on AI tools to conduct hacking tasks, according to Cybernews. The group has listed prominent organisations among its alleged victims, including Visa, Mastercard, Walmart, JPMorgan and Starbucks, according to Cybernews.

What makes this dump different from a routine leak-site claim is that researchers read the files. Cybernews researchers examined the exposed data and described its contents in detail, and re-reports from Enterprise IT World and CryptoCompass repeat the same breakdown. The contents are the story, not the ransom demand.

Three datasets, three different years

The dump is not one breach snapshot. It is three datasets from three different years, and that mix is the first clue about where the data came from. The table below is built from Cybernews’s published counts.

Dataset Records Date range What it contains
Smart TV viewing events About 150,000 2024 Timestamps, viewing durations, device information, city, ZIP code, TV service provider. No programme titles.
Political advertising About 42,600 September to October 2026 Political affiliations, issues discussed, sponsors, estimated advertising costs, timestamps, plus one political advertisement video.
AI advertising model training About 99,100 2021 Embedding vectors, mathematical representations of training data.
Total, by our calculation About 291,700 2021 to 2026 Three datasets spanning five years.

By our calculation, the three datasets add up to about 291,700 records, 150,000 plus 42,600 plus 99,100. A single 220MB dump spanning five years of Samsung advertising data fits a vendor-side advertising-data store far better than a live breach of Samsung’s own systems. That is consistent with, but not proof of, Samsung’s third-party-vendor explanation, and the article keeps it that way.

The largest dataset holds around 150,000 Samsung smart TV viewing events recorded in 2024, including timestamps, viewing durations, device information, cities, ZIP codes and TV service provider names, according to Cybernews. The viewing-events dataset does not expose what individual users watched, because no programme titles are included, according to Cybernews, corroborated by reef-sentinel.net. It shows viewing patterns and usage data collected from Samsung devices.

The second dataset contains approximately 42,600 records related to political advertising aired between September and October 2026, including one political advertisement video, according to Cybernews. The records show political affiliations, issues discussed in advertisements, sponsors, estimated advertising costs and timestamps. The third dataset holds around 99,100 records associated with AI advertising model training from 2021, primarily embedding vectors, according to Cybernews.

The exposed files also included two internal documents, according to Cybernews. One, dated September 2026, outlines an agreement governing the proper use of collected data and contains the full name and email address of a Samsung employee. The other, an access questionnaire for what appears to be Samsung’s Content Data Warehouse, reveals information about an individual employee and details about the company’s infrastructure.

Most of the leaked material is old. The viewing data is from 2024 and the AI training records from 2021. Only the political advertising records, from September and October 2026, are recent. The records mostly relate to US TV audiences, according to reef-sentinel.net, and it is not yet clear how many unique households or devices are in the data, according to the same source.

What Samsung confirmed, and what it did not

Samsung confirmed the incident to Cybernews on 20 November 2026, the day after Cybernews first reported the dump and six days after n0n listed the company. A Samsung spokesperson said the company is aware of the cyber incident and that it involves a third-party vendor that supported advertising data processing for Samsung’s North America Service Business, known as Samsung Ads, according to Cybernews. The spokesperson’s full statement reads: “The vendor’s preliminary investigation indicates that some Samsung information may have been affected in the incident. We are working with the vendor on an investigation to determine the scope of the incident. We will share additional information as our investigation progresses. The protection of our customers’ and partners’ information is a top priority for Samsung.”

That statement confirms a third-party vendor incident. It does not confirm that n0n stole data from Samsung Electronics directly, and it does not confirm that the 220MB dump came from that vendor. Samsung did not name the vendor, and the scope is still being determined, in the company’s own words. The two halves of this story, the confirmed vendor incident and the claimed n0n dump, have to stay separate until Samsung or an investigator connects them.

The dump’s origin is not established. The data could come from Samsung’s own systems, from the third-party advertising vendor Samsung named, from an advertising or analytics partner, or be a mix of older material, according to Cybernews and reef-sentinel.net. Whether any Samsung customer names, email addresses or passwords were in the dump is not established. The viewing records as described contain device information, cities, ZIP codes and TV service provider names, and no customer names have been reported in any source read.

Some re-reports published on 20 November 2026 still state that Samsung had not commented or had yet to comment. That wording is outdated, because it predates or missed the Cybernews update of 20 November carrying Samsung’s statement. No tier-1 outlet, including Reuters, AP, BBC, BleepingComputer, TechCrunch, The Verge or The Register, had covered the story as of 20 November 2026, and Cybernews remains the only substantive source.

The leak is the evidence about your TV

The most newsworthy finding is not the ransom claim. It is what the leaked files appear to show, namely that Samsung smart TVs collect viewing data and send it to Samsung’s data centers, according to Cybernews. The consumer question is therefore not whether your TV was hacked, because nothing in the leak suggests the televisions themselves were remotely compromised, according to reef-sentinel.net. The data looks like back-end analytics, not something taken from devices. The question is what viewing data your TV sends, and whether you can stop it.

Samsung is the world’s biggest TV brand by revenue share, with 29.1 percent of the global market in 2025, according to Omdia data reported by TechNext24. That scale is why a vendor-side advertising leak matters even without a confirmed count of affected users. No affected-user count exists, because 150,000 viewing events is not 150,000 users, and no source has established how many unique households or devices are in the data.

The other Samsung breach this month is a different story

Do not confuse this leak with the other Samsung Group incident in November. Multicampus, a Samsung Group education affiliate and subsidiary of Samsung SDS that runs corporate employee training and the OPIc English speaking test in Korea, suffered unauthorised external access to the cloud environment used for its Learning Platform online training service on 9 November 2026, according to The Asia Business Daily. The company notified its client companies and reported the incident to South Korea’s Personal Information Protection Commission on the night of 11 November.

The Multicampus breach is believed to have exposed the personal information of approximately 70,000 Learning Platform users, including names, phone numbers, email addresses, department names, employee numbers, IDs, ranks and positions, according to The Asia Business Daily and Seoul Economic Daily. Resident registration numbers, Korea’s national ID numbers, were not compromised, and no financial damage has been reported. There is no sign that the Multicampus breach and the n0n Samsung claim are connected, according to reef-sentinel.net. They are two separate November incidents involving different Samsung Group entities.

Two older stories also surface in searches for Samsung data breaches and must not be mistaken for this one. In 2025, roughly 270,000 of Samsung Germany’s customer support tickets were dumped online, exposing names, emails, home addresses and order details, according to Cybernews. Cybersecurity firm Hudson Rock linked the breach to credentials stolen in 2021 that had not been rotated. The 2022 Lapsus$ Samsung breach, about 190GB of source code, is a different, four-year-old incident. A 26 October 2026 Cybernews story about Samsung smart TVs allegedly collecting audio telemetry is a separate, older story about alleged TV spying, not a data breach, and not this incident.

What should I check on my Samsung TV?

Check your Samsung TV’s privacy settings first, because the leak shows your TV sends viewing data to Samsung’s data centres. Turn off Viewing Information Services and Interest-Based Advertising in the privacy menu, and use the Do Not Sell or Share option if your set has one.

  1. Open your Samsung TV’s privacy settings and find the Viewing Information Services toggle to switch off.
  2. Turn off Interest-Based Advertising in the same privacy menu so your viewing data stops targeting ads.
  3. Use the Do Not Sell or Share My Personal Information option if your TV shows one.
  4. Check the setting again after a software update, since menu names and defaults can change.
  5. Watch for phishing messages that quote your viewing habits, and report suspicious ones to Samsung directly.

The toggle lives in the same place across recent models, according to Consumer Reports. Press Home on the remote, go to Menu, then Settings, then All Settings, then General and Privacy, then Terms and Privacy, where Viewing Information Services and Interest-Based Advertising sit. Tom’s Guide gives a shorter route on some models, going through Support and then Terms and Privacy, with Privacy Choices as the entry point. Menu names vary by model and year, and older sets may put the same choices under Support, then Terms and Privacy, so look for the closest match. Samsung has said before that its Automatic Content Recognition viewing-data feature is opt-in and can be turned off, according to reef-sentinel.net.

What to watch next

Whether Samsung names the vendor, and whether the scope of the incident is confirmed. Whether a tier-1 outlet picks the story up, which would be the first independent confirmation beyond Cybernews. Whether investigators establish that the 220MB dump came from the vendor Samsung named, from Samsung’s own systems, or from somewhere else entirely. And whether Samsung says anything about how many households or devices sit behind the 150,000 viewing events. Until then, the confirmed facts are a vendor incident and a dump of advertising data spanning five years, and the claimed facts stay claimed.