Hacker News new | past | comments | ask | show | jobs | submit login

> Is there any reason not to assume that the Chinese government is surveilling Zoom calls en masse?

It makes perfect sense to boycott Zoom based on their security issues, but does the presumption of innocence mean nothing anymore? It shouldn't be difficult to prove that Zoom is actively trying to block Chinese activists.

China banned Zoom during the trade war. Are you going to treat that as evidence that Zoom isn't colluding with the CCP? Why is Zoom not embracing end-to-end encryption to free users evidence that they are beholden to China, but Zoom committing to end-to-end encryption for paid users not evidence of the contrary? This type of circumstantial cherry picking is how conspiracy theories sprout.




Is it often that China unblocks a communication platform without them fully complying with the CCP’s desire to monitor and censor everything?

I can’t think of any major platforms that got banned and reinstated with zero changes. I’m treating that as evidence that they are colluding.


For some reason, China periodically bans and unbans websites before banning it completely. Reddit is one example of this. It looks like Zoom is moving away from the Chinese market altogether:

https://technode.com/2020/05/15/zoom-suspends-chinese-indivi...


does the presumption of innocence mean nothing anymore

In a court of law, yes. In the court of public opinion, it has never been that way.

I don't blame people for being suspicious of Zoom. Why wait until the harm has already been done to move to something else?

Zoom knows it has a credibility problem. If Zoom really cared, it would do something to distance itself from China and the Chinese government. But it doesn't.


I don't blame people for being suspicious of Zoom either. I just find it sad that so many readers of Hackernews are endorsing logic that is so unscientific.


In what way is it "unscientific"? That's an odd choice of words in this situation. What we're aiming for is rational and in this instance, it's perfectly rational to be suspicious based on an abductive argument and unproven extrapolation. As another commenter aptly noticed, once you're able to prove it, the harm has already been done.


Having read the original comment again, I can see how it can be interpreted as suggesting to avoid Zoom as a precautionary measure, rather than accusing Zoom of wrongdoing. However, plenty of people are treating this incident as proof that Zoom is lying about encryption.


Don't want to sound I'm defending Zoom or CCP, but how does en masse surveillance work technically in this case? How practical it is to go over all the voice data (not manually I presume)? Text based mining and censoring is possible but there still is a huge gap and language expertise isn't cheap. I might be missing a few references but it'll be great if someone could point me to these links.


I might be wrong here, since I only have this from hearsay, and I certainly won't touch Zoom. But, wasn't the "end-to-end encryption" a term they used in a blatantly deceptive way? The two ends not being the two users, but the user A and zoom servers, and zoom servers and user B? Please do correct me if I'm misinformed.

Edit: I checked. They've got you fooled.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: