What TV Manufacturers Won't Tell You About Your Viewing Data
I sent 17 questions to every major TV manufacturer about how they track your viewing. Most never answered. The ones that did exposed practices you won't find in any privacy policy.
After LG got accused of spying through its TVs, I wanted to see if the rest of the industry was collecting and protecting data any differently. It isn't. Every TV manufacturer is gathering information on what you watch, and while some of it is necessary for the hardware to function, much of it isn't. It includes details about your viewing that you'd probably prefer stayed private.
So I asked them directly what they're doing with all that data. The responses range from silence to answers so carefully worded they say almost nothing. And that gap between the official policy and the actual practice is where your habits are really being stored.
The 17 Questions
We sent each manufacturer 17 specific questions about data collection, retention periods, and third-party sharing, because privacy policies are drafted to be unreadable and answer every inquiry with the same vague reassurance. When you ask "exactly how many days do you store voice recordings?" instead of "do you respect privacy?", you get responses you can actually compare across brands. My take is this direct approach is worth the effort: a policy stating "we may share data with partners" tells you nothing actionable, but a specific answer about whether microphone audio is retained indefinitely or deleted after processing lets you decide if the feature is worth the trade.
The questions targeted retention and third-party sharing because those details determine what happens to your data after collection. Samsung's response confirms the embedded microphone only listens locally for the wake word "Hi, Bixby" and ignores private conversations, but it stops short of stating a retention period for that audio. TCL acknowledged that on some models, third-party operating systems handle microphone data, which means Google's or Amazon's privacy terms apply instead of TCL's. Compared to generic policy language about "trusted partners," these answers at least identify the specific companies involved and admit that voice data leaves the device.
The limitation is that brands still route these answers through legal teams, so you get technically responsive but strategically vague replies. Several manufacturers told us data is kept "for as long as necessary," which is a retention period in the same way "a while" is a time estimate. The 17 questions force specificity, but they don't force a number.
question_categories = {
"voice_data": "Retention period for microphone recordings (in days)",
"third_parties": "Specific companies receiving viewing or voice data",
"deletion": "User-accessible method to purge all collected data"
}
samsung_response = {
"microphone_scope": "Local wake word detection only",
"retention_days": None # Not specified in response
}How Your TV Watches You Back
Your TV knows what you're watching because it's listening to the show, not you. Automatic Content Recognition works by sampling audio from whatever's playing, generating a fingerprint, and matching it against a database of content. Samsung, LG, Vizio, and TCL ship this enabled by default on most models from the last decade. The opt-in is buried three menus deep under Terms & Privacy, and even then, disabling it doesn't always stop the data collection—it just stops the personalized ads.
The microphones add another layer. Samsung says its TVs only listen for the wake word "Hi Bixby" and process that locally, but the hardware stays powered. TCL goes further: their remotes and TVs have microphones that third-party apps like Google Assistant can access. Your voice commands end up on servers you didn't explicitly authorize, and the mic remains active even when you're not using voice features.
Here's what the audio fingerprinting looks like in practice:
import hashlib
def generate_fingerprint(audio_chunk):
# TV samples audio every few seconds, hashes it, and sends to server
return hashlib.sha256(audio_chunk).hexdigest()[:16]
fingerprint = generate_fingerprint(b"audio_data_from_speaker")
print(f"Content ID: {fingerprint}")My take is that ACR is worth disabling if you care about privacy, but expect friction. On Samsung TVs you can turn it off at Settings > Support > Terms & Privacy > Viewing Information Services, yet the limitation is that you can't verify the hardware is truly inert without physically disconnecting the microphone or network. Manufacturers don't provide audit logs, so you're trusting the toggle switch.
Compared to a dumb TV from 2010, which collected zero viewing data, a modern smart TV treats every show you watch as a data point. That's the difference between owning a television and renting a Nielsen box that lives in your living room.
Who Answered and Who Didn't
I think the pattern here—regulators structuring inquiries so companies can technically comply without actually answering—represents a real erosion of accountability, not just a PR problem. When Apple can respond to pointed questions with indirect deflections and still check the box of "cooperation," we're not seeing transparency; we're seeing the performance of transparency. The frustration about the article obscuring those indirect answers hits on something important: when journalists accept corporate non-answers as answers, they launder the evasion. I don't think this is malicious most of the time—it's just easier to write "Apple declined to comment" than to parse the careful language of a response that technically says nothing. But the result is the same: readers walk away thinking they know what happened when they don't. What's less discussed is how the unusable website factors in. If the primary source material is buried behind a broken or hostile interface, the "answer" becomes effectively inaccessible to anyone without dedicated time to fight the UX. That's not an accident of design; it's another layer of friction that serves the company's interest in making scrutiny costly. One hardware gotcha that rarely gets named: the power button on most smart TVs doesn't actually turn the set off. It drops the device into standby mode, which keeps the network stack alive and continues logging activity even when the screen is black, so the intuitive move of "just turning it off" at night doesn't create the air gap people assume it does. The question worth sitting with: if every avenue for getting a straight answer—regulatory, journalistic, technical—has been smoothed into something that looks like engagement but functions like evasion, what actually forces a company to say something true?
Who Answered and Who Didn't
Conclusion
The answers that came back paint a picture of an industry that treats your living room like a data center. We asked 17 questions, and while the top three manufacturers and streaming device makers responded, the gaps are telling—and the Vizio case is worse than a gap. When Northeastern researchers found Vizio kept transmitting after users opted out, the company’s response that it was just “routine operational data” doesn’t hold up when you consider what that data actually contains. Opting out is supposed to be the escape hatch, but right now it looks more like a suggestion. I’m still not sure what to make of the silence from the companies that didn’t respond. Maybe they’re embarrassed, maybe they don’t care, maybe both. There's also a technical leak in the obvious fix: HDMI-CEC handshake data can still tell the TV what you're watching even when you're using an external device, so that "dumb" screen isn't as blind as it looks. What I do know is that until regulators force the issue or consumers start treating smart TVs like the surveillance devices they are, the default will remain: your viewing habits are inventory. My move today is to disconnect the TV from the internet entirely and use a separate streaming device I control—or accept that every show I watch is also watching me back.