Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

South Korean regulator accuses DeepSeek of sharing user data with ByteDance

bbc.com

Post reply on HN
Print view View original

61–70 of 146 posts

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#61

Post by HeatrayEnjoyer »

nicce wrote: post #58
varenc wrote: post #31

Some missing context is that the data is shared via the DeepSeek app's use of ByteDance analytics/configuration frameworks. So not a backroom deal where DeekSeek handed over the chat history for its user base, but rather ongoing analytics data being sent from the DeepSeek mobile app. Here's the SecurityScoreCard article that brought attention to this: https://securityscorecard.com/blog/a-deep-peek-at-deepseek/#... Be…

So Deepkseek is not sharing more data than most advertising-funded apps in the world?

Only if they were breaking the law too.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#62

Post by bilbo0s »

unclebucknasty wrote: post #49
varenc wrote: post #31

Some missing context is that the data is shared via the DeepSeek app's use of ByteDance analytics/configuration frameworks. So not a backroom deal where DeekSeek handed over the chat history for its user base, but rather ongoing analytics data being sent from the DeepSeek mobile app. Here's the SecurityScoreCard article that brought attention to this: https://securityscorecard.com/blog/a-deep-peek-at-deepseek/#... Be…

Interesting, but I don't think those details will be ameliorative to the people who are concerned (e.g. U.S. Congress). In fact, I wonder if it may further underscore their concerns, given that it surfaces the interconnectedness between all of these firms.

Yep.

No one cares about the details. (Heck, I'd be willing to wager good money that the politicians and most of their staffers don't even understand the details). In the end, it's just one more reason that Chinese models will not be legal in the US in the near future.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#63

Post by itishappy »

kelnos wrote: post #57
itishappy wrote: post #53

Earlier quoted context omitted.

com·plain /kəmˈplān/ express dissatisfaction or annoyance about something

Nothing in the sentence you quoted suggests that anyone is complaining about anything.

We're leveling allegations, but without dissatisfaction or annoyance?

"Hey you, you're sharing our data! No further comments. Carry on!"

No, an allegation is a claim that something is wrong.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#64

Post by blackeyeblitzar »

gruez wrote: post #60
blackeyeblitzar wrote: post #39

Earlier quoted context omitted.

That won’t protect you from its propaganda/censorship. Some versions of DeepSeek’s models have bias built in - as in it’s not just implemented by their service/app. But offline does protect you from privacy/security issues.

>Some versions of DeepSeek’s models have bias built in Is this empirically supported? IME it doesn't hesitate to talk about Tienanmen Square or whatever, so if there's "bias" it must be very well hidden.

Yes if you search, lots of people have shared evidence of this. But it depends on which model you’re using, as some seemingly don’t have the bias built into their training.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#65

Post by gruez »

SecretDreams wrote: post #40
Sohcahtoa82 wrote: post #8

This is my surprised face -_- If you're shocked or even the slightest bit surprised, then I can't imagine how blissful your life is to be so unaware about how much corporations are sharing data with each other. Like, I wholeheartedly expect that if I mention Beyblade toys on Facebook, then the next time I visit Amazon, they'll be suggesting Beyblades even if I've never even searched Amazon for toys, let alone Beyblad…

You could literally talk about beyblade toys on a whatsapp video call and you'll be getting Amazon ads for dem blades the next day.

There's basically no credible evidence of this happening. All there is are vague anecdotes which are easily explained with confirmation bias and/or the birthday paradox.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#66

Post by ivewonyoung »

bilbo0s wrote: post #62
unclebucknasty wrote: post #49

Earlier quoted context omitted.

Interesting, but I don't think those details will be ameliorative to the people who are concerned (e.g. U.S. Congress). In fact, I wonder if it may further underscore their concerns, given that it surfaces the interconnectedness between all of these firms.

Yep. No one cares about the details. (Heck, I'd be willing to wager good money that the politicians and most of their staffers don't even understand the details). In the end, it's just one more reason that Chinese models will not be legal in the US in the near future.

> it's just one more reason that Chinese models will not be legal in the US in the near future

This isn't about the model, it's about the mobile app.

The open source model weights are different from the website and the app. The model cannot track you.

Not just Congress, even techies can be confused about these things.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#67

Post by gruez »

blackeyeblitzar wrote: post #36

Reminder on various DeepSeek problems: 1. DeepSeek is full of propaganda/censorship ( https://arstechnica.com/ai/2025/01/the-questions-the-chinese... ) 2. They already had a serious security and privacy issue when they left their database wide open and leaked everyone’s chat history ( https://www.wiz.io/blog/wiz-research-uncovers-exposed-deepse... ) 3. Multiple teams of security researchers found code that links Deep…

>1. DeepSeek is full of propaganda/censorship (https://arstechnica.com/ai/2025/01/the-questions-the-chinese...)

Your source doesn't even mention "propaganda". Moreover, while censorship is concerning is concerning, I don't see how that practically affects users. If I want to know how to center a div, who cares if it's cagey about what happened in 1989?

"Western" AI is also arguably full of "propaganda/censorship". Remember when chatgpt just came out, and conservatives were lambasting it for being "woke"?

>3. Multiple teams of security researchers found code that links DeepSeek to the Chinese government through China Mobile, who is banned from operating in the US (https://www.pbs.org/newshour/world/researchers-link-deepseek...)

Seems like a nothingburger?

"Neither Feroot nor the other researchers observed data transferred to China Mobile when testing logins in North America, but they could not rule out that data for some users was being transferred to the Chinese telecom."

>4. Other countries like South Korea are banning DeepSeek already over privacy concerns (https://mashable.com/article/south-korea-blocks-deepseek)

That's what this thread is about. Many commenters have mentioned why South Korea's actions were dumb.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#68

Post by nickthegreek »

blackeyeblitzar wrote: post #64
gruez wrote: post #60

Earlier quoted context omitted.

>Some versions of DeepSeek’s models have bias built in Is this empirically supported? IME it doesn't hesitate to talk about Tienanmen Square or whatever, so if there's "bias" it must be very well hidden.

Yes if you search, lots of people have shared evidence of this. But it depends on which model you’re using, as some seemingly don’t have the bias built into their training.

And they also shared how trivial it is to bypass. It is one of the most uncensored released models that you can run privately from my understanding.

If you have a better model for the OP to run, please present it.

Edit: https://huggingface.co/perplexity-ai/r1-1776

Just released.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#69

Post by nomel »

yalogin wrote: post #51

Shouldn't this be the other way around? TikTok has the most user data for any LLM to train with. I bet they will make a killing with it, unless of course the CCP decrees that they share it for free.

What useful textual user data do you see coming from TikTok? All the text seems very low quality, to the point where I naively assume that including it in training data would decrease performance.

Re: South Korean regulator accuses DeepSeek of sharing user data with ByteDance

#70

Post by pepinator »

ivewonyoung wrote: post #66
bilbo0s wrote: post #62

Earlier quoted context omitted.

Yep. No one cares about the details. (Heck, I'd be willing to wager good money that the politicians and most of their staffers don't even understand the details). In the end, it's just one more reason that Chinese models will not be legal in the US in the near future.

> it's just one more reason that Chinese models will not be legal in the US in the near future This isn't about the model, it's about the mobile app. The open source model weights are different from the website and the app. The model cannot track you. Not just Congress, even techies can be confused about these things.

Yes, exactly, what the guy above was saying is that they're just looking for excuses to keep people from using the Chinese thing.
Post reply on HN