The fact that they show end users (no pun intended) an "end-to-end encrypted" badge on the meeting window itself, and elsewhere explain how a Zoom server (not Zoom client) is what constitutes an "end" despite the whole rest of the electronic communication industry using "end-to-end" to refer exclusively to user agents, is bonkers.
Agreed. It's unlikely they stumbled onto an industry standard phrase like that alone, then innocently used it without knowing the generally accepted meaning. This is deceptive advertising.
This bothers me so much - as I have personally recommended zoom to many people and customers. People that have likely seen that I spend an unusual amount of time focused on computer security, backups, and care in communications.
If I remember correctly, zoom used to have on the front page - use this it's encrypted, and 'even used by us govt something-something' - so I assumed it was completely secure.
I actually refused to use other conference services much to the bemoans of many clients who already had other 'goto' software installed, used and understood - and convinced them that in order to talk turkey we needed to use the real secure zoom system.
It passed the smell test at the time for me because they also had paid plans which meant to me a legitimate business that did not need to slay privacy with ads and such, as they had a clear path to make money.
Now they are tarnished, and my reputation with several clients and doctors and others may be as well - as this is getting mainstream press (I think that's a good thing actually) - I'm livid about this.
I agree with below it is also fraud - and another commentator mentioned they changed to "your client connection is encrypted" is still deception imho.
Needs a big asterisk and real explanation of the lack of privacy.
I doubt non-technical users are savvy enough to find out about this problem and they probably won't remember to blame you for something clearly beyond your control.
I know it's the kind of thing that can randomly keep a person up at night, but I think you can probably safely forget about this awkwardness and move on.
Thing is - this story is all over the mainstream news. It's got multiple stories on the yahoo front page, it's reported via multiple sources in my fbook feed, but also with our regular tv news station.
People are seeing this about zoom that have never seen zoom.
People that I suggested use zoom are going to notice the word even if they would normally scroll past some gomeeting story.
any anyone who is within earshot of such 'non-tech news' is hearing how unsecure zoom is.
Sure most of my clients are unlikely to read HN at all, and most are unlikely to read tech crunch regularly if at all - but I bet some have TC or something similar in their fbook feed.
However people who don't even own computers are seeing this debacle.
So, anyone I've advised to use zoom for privacy and security, citing the encryption and use by US gov - is going to have to wonder - how do these things happen on a secure, private, encrypted system - must not be what it was purported to be by that guy Steve. Then they are going to wonder what kind of damage could be done with the info that was 'securely' shared with the service.
Because all the others I have been inspecting failed one of the big checkboxes almost immediately. Is it private? most no - skip. Is it easy to get started? the private private self hosted, not so much..
Zoom - private, encrypted, used by the us govt and easy to use.. So others trust and use, they have a firm business model - and a reputation not to lose. No others I found had this.
Sadly, my experience has been the opposite, because open-source stuff usually takes more work to run and is seldom as polished, and users hate that. I rarely, if ever, recommend open-source anything for end-user use. Zoom did well because it had the least hassle of any solution, hands-down. I'm not saying it's impossible to subscribe to a hosted open-source service that's as good, but it doesn't exist yet.
"Fraudulent", at least as far as I know, is reserved for "intentional deception". It could be that Zoom is indeed doing this intentionally, but without proof of it being intentional, I don't think we should assume so.
The IT guys in the industry know very well what constitutes an E2E encryption. Those two ends must be "trusted" which means it's either you yourself - your computer, or the other party which you want to talk to. Everything in between is third party and must get only encrypted data. If they redefine one of the "ends" as Zoom server, that's definitely intentional, blatant, and therefore fraudulent.
I'm not disagreeing that they are using E2E wrong, I fully agree with you there.
But just because someone uses a word wrong doesn't give you any proof about their intentions. See https://news.ycombinator.com/item?id=22767447 for further elaboration on that point.
Again, it's harmful to use words incorrectly, _especially_ when it comes to E2E, so they should rightly get flak for getting it wrong. You all seem to be so sure that it was intentional though, while I've seen the same problem so many times before in the industry without it being intentional. If you do have proof it's intentional, please share it with the rest of us so we can be on the same page.
Eh, haven't large companies leaked private details from their customers time and time again, in basic ways like forgetting that they have backups on S3 buckets with zero protection?
Being incompetent has nothing to do with the size or prominence of the company. Big/prominent companies fuck up/are sloppy all the time.
This is not a fuck up or negligence, Zoom deliberately used "end-to-end" with a completely different meaning than the rest of the world. I can't describe this as anything else than malicious and fraudulent.
Their intention was to deceive users that the communication was encrypted, when in reality it wasn't.
I guess you don't have experience working in a company with a marketing team that feel they can buzzword things in without checking with IT teams then.
I'm in no way saying it's impossible that Zoom did say E2E encryption while knowing that's not true, but I could imagine a scenario where a security person says "Yeah, we're encrypting connections to our backend" and a marketing person researching E2E and then saying to themselves "Yeah, sounds like we're doing E2E, let's write that", because this stuff happens all the time in the industry.
> Their intention was to deceive users
You sound so sure about their intentions, do you have any actual proof of this that others are missing? Again, I'm not saying it's impossible that their intention was to deceive users, but as an engineer, I always favor proof over guessing.
Not defending zoom here -- they done fucked up -- but there is a huge disconnect between the marketing folks and the technical folks. It's possible that E2E Encryption was something they planned on implementing but haven't, and the marketing department either didn't get the memo or didn't understand and still kept the wording.
If marketing misuses a term on the website because they didn't understand what it meant, that's their fault. If they listed a feature that hasn't yet shipped, that's their fault.
It's natural that there is a divide and marketing isn't expected to understand every engineering thing (nor the other way around.) If your job is to write words, though, you are responsible for the words you right.
That's not an excuse. Sooner or later it'll become a faux pas to mis-use basic security terminology. Obviously most of the public doesn't care about most of the terminology. But "end to end" has been trending for a decade at least.
I don't believe that's a thing. Thing like this will run over a lawyer's desk to make sure they're covered and won't get sued for billions of dollars, and they will make sure that it's legally sound, so they will check whether they have it, and they won't rely on "it's a planned feature maybe some day we will have it".
> To be clear, in a meeting where all of the participants are using Zoom clients, and the meeting is not being recorded, we encrypt all video, audio, screen sharing, and chat content at the sending client, and do not decrypt it at any point before it reaches the receiving clients.
The first problem is:
> Zoom currently maintains the key management system for these systems in the cloud.
Obviously, this compromises many of the benefits of e2e encryption. Having said that, it doesn't remove all of the benefits, and it's a (bad) precedent that has been set by other companies (eg. apple) where keys for end-to-end encrypted communication are backed up to the cloud.
The second problem is that Zoom has a second class of "client" called a "Connector" which runs in the cloud, and also has access to the keys for decrypting the stream. I definitely think that when one of these connectors is being used, it is false advertising to show the "e2e encrypted" status. However, there are clear technical reasons why these connectors are needed. Being able to dial into a meeting from an ordinary phone is important functionality that simply cannot support end-to-end encryption.
The interesting section to me is the later paragraph:
> For those who want additional control of their keys, an on-premise solution exists today for the entire meeting infrastructure, and a solution will be available later this year to allow organizations to leverage Zoom’s cloud infrastructure but host the key management system within their environment. Additionally, enterprise customers have the option to run certain versions of our connectors within their own data centers if they would like to manage the decryption and translation process themselves.
In particular, being able to use your own key management system would make this truly end-to-end encrypted by any definition, even if you are still using Zoom's cloud infrastructure.
> other companies (eg. apple) where keys for end-to-end encrypted communication are backed up to the cloud
Backed up for iCloud users who might not know any better, but not backed up for people who take the time to learn how to guarantee the full protection of E2E by keeping iCloud off. The fact that the full benefit is available with little effort, albeit not obvious, creates a contrast to how:
> Zoom has never built a mechanism to decrypt live meetings for lawful intercept purposes
...but they easily could. Users can't just search for how to harden the Zoom encryption to the point of lawful intercept becoming impossible and find a simple solution the way they can with Apple.
> an on-premise solution exists today for the entire meeting infrastructure
...is not practical for most.
So, eliminating the E2E badge was the right move. The fact that it was there until now is shady.
It means that it's opaque to your ISP and other people on your network, which is at the very least useful for people using it at McDonald's or a university.
So, pretty much the opposite of all those people currently using it instead of going together to McDonald's or university?
On a more serious note, you would need to trust not one ISP if the video wasn't encrypted, but all the ISPs in the room simultaneously, and that is probably something even less trustworthy than Facebook.
They are using the same language as Apple talking about iMessage and FaceTime. Apple talks about end to end encryption, but one end is iCloud which is why you can get your messages simultaneously on all devices.
There is a big difference there though. In transit, iMessage and FaceTime backups are end-to-end encrypted, it's just the iMessage backups on iCloud that also store the key.
FaceTime chats, though, truly are end-to-end encrypted and the calls aren't backed up like iMessages are.
They merely mention that backups to iCloud happen automatically by default, and not that doing so means the default is that Apple can view and decrypt all your messages.
That's not true at all. iMessage is really end-to-end encrypted.
Two options impact it: "Messages in iCloud" re-encrypts and uploads messages to the user's iCloud account and stores the key in iCloud Keychain (also end-to-end encrypted).
Only when enabling iCloud backup will that key be revealed to Apple.
Yeah, the parent conversation was about whether iMessage is end-to-end encrypted. GP claimed that it cannot be due to it supporting multi-device sync, which is not true.
The quiet part is "American citizens of Chinese descent can't be trusted". This is as obvious of a dogwhistle as when people "innocently" point out that the CEO of some maligned Wall Street firm is Jewish and I have no idea why it's getting upvotes on this site.
This is not true. Don't spread FUD. Apple does not have the ability to read your messages. All messages stored on their servers are encrypted with keys that live only on the phone.
iMessage doesn't store your decryption keys on Apple's servers unless you opt into iCloud backup which is a whole different service and security concern.
Most people use iCloud backup. Even if you don't, your messages are still sent to Apple by the recipient. And Apple prohibits third party backup services.
> Apple does not have the ability to read your messages.
iCloud backup is an Apple service and it has the ability to read most of your messages even if you don't use it, which makes this statement categorically false.
This is completely ridiculous. iMessage is encrypted by my device and remains encrypted until it gets to the recipient device. That is what end-to-end encryption means.
That I may have given Apple my private key through a different message in no way affects that end-to-end encryption, because it is trivial to decide not to give Apple that key.
iCloud isn't some separate entity from iMessage. It's all Apple. And you have no option to use a different cloud backup provider.
You can decide not to give your keys to Apple, but you can't decide for all your friends to not give their keys to Apple, and the result is the same: Apple can read your messages.
And the marketing is so misleading that hardly anyone knows that Apple can read most iMessages.
Sorry, let's be explicit here, as you seem intent on muddying the issue. Where, other than the endpoints, is the message decrypted when people use iMessage? Your succinct answer to that will clear this up for everyone.
On GCBD's servers in China. Possibly on Apple's servers in the US if they are running a wiretap. Due to the way key distribution works for iMessage, it is trivial for Apple and GCBD to do so.
Your message, through several layers of indirection, relies on a security conference paper from 7 years ago[0] + the assumption that Apple haven't updated the protocol in 7 those years.
No, my message relies on the fact that people have been looking at iMessage for years, and nobody, least of all Apple, has said that the implementation changed in any way to prevent Apple from viewing the messages.
Here is another article from 2016, which shows that Apple patched iMessage to prevent attackers who don't have access to Apple's servers from reading the messages but still kept the ability to read the messages themselves. https://blog.cryptographyengineering.com/category/imessage/
Apple was aware that people knew it could decrypt iMessage messages this entire time, but Apple made no changes that would fix that. That should give you some idea of whether Apple intends to ever fix that.
Apple can, of course, do whatever it likes, up to simply recording the screen and sending that to weird & wonderful government agencies. Like almost everything in mainstream security, it comes down to who you trust. It doesn't mean it isn't E2E though.
E2E encryption simply means that messages are only decrypted at the endpoints. That certainly isn't true of iMessage in China, and it might not even be true for some users in the US — we have no way of knowing because the protocol makes no guarantee against it.
So basically the first and second parties themselves need to do all encryption and decryption without any help from the third party running the service. Which is the age old usability issue famously holding back the casual adoption of PGP. Hard enough with text... To do it with video conferencing would be quite the feat. Someday, though.
I have linked it several times in this thread. Here it is again:
"If you have iCloud Backup turned on, your backup includes a copy of the key protecting your Messages. This ensures you can recover your Messages if you lose access to iCloud Keychain and your trusted devices."
Only Apple can know exactly when or where or how often they decrypt people's messages from their backups, because once they have the keys they have the means to do it at any place and time, for any reason, without anyone's knowledge or consent.
What we know is that they can and do decrypt iMessages from iCloud backups in response to law enforcement requests[1]. This proves that they hold the keys, if their own support pages weren't enough evidence for you.
And even if none of that were the case, couldn't they just push out an update to the app or OS (just to the target, so other researchers debugging or watching traffic wouldn't know) which would cause the device to exfiltrate the cleartext anyway? Or always have had said feature?
"If you have iCloud Backup turned on, your backup includes a copy of the key protecting your Messages. This ensures you can recover your Messages if you lose access to iCloud Keychain and your trusted devices."
Let's not use sarcasm or sources..... Let's puzzle it out.
You don't use a password to encrypt your iCloud backups... They're specific to the hardware your backing up. If you have an itouch for example it's backups are separate from your phone.
So now you have these backups in the cloud and you lose your iPhone, you remote wipe it.
Now your new one arrives and you restore from backup... Your iMessage private keys are available to apple unencrypted .... Because you didn't need to provide a second factor of authentication for unlocking the backup you were just asked which one to use.
Apple and any reputable nation-state can read your iMessages with a subpoena ... If you use iCloud backups and not local backups with a password.
2) What about your iCloud account and password that are required to encrypt, store, access, and decrypt the backups there? Is that not a factor worth consideration?
Your password is not a factor worth considering. You can ask Apple to change it. That means they have the ability to change it. That means they have access.
This is both true and false. Apple stores keys on the device so they can't read your old messages, but say they want to start reading messages of a particular user, they can simply issue a new key and store it on the device and the server and start decrypting the new messages using it.
This is why WhatsApp for example notifies users when the key of the recipient changes, and they give you a way of verifying that the both keys at both ends are identical.
Tuxer said "keys," not "your decryption keys." Apple distributes the public keys that each party encrypts their message with, and they route the encrypted messages through their servers. They can trivially eavesdrop on conversations by simply providing a key from a key pair they generate to a participant and reencrypting messages using the other parties' public keys after deciphering the messages.
Yes, it does. The messages are 'end to end' encrypted in the iMessage service, but then iMessage backs up its encryption key in the iCloud backup service, defeating the point.
"If you have iCloud Backup turned on, your backup includes a copy of the key protecting your Messages. This ensures you can recover your Messages if you lose access to iCloud Keychain and your trusted devices."
That is true of any end-to-end solution. If you back up your private keys, anyone who has access to your backup would be able to access the encrypted messages. Remember, you can turn off iCloud backup if you're worried about Apple accessing your keys.
Ultimately, it's false to equate iMessage's encryption scheme, which is end-to-end, to an encryption scheme that requires a server to relay decrypted data.
Utterly false. Real end-to-end encryption would encrypt the backup with a key that is not available to the backup service (e.g. derived from a passphrase not sent to the server).
Of course this system has better usability, which is why Apple does it. But it's still a farce to call a system where Apple has the ability to decrypt the majority of messages "end-to-end" encrypted. The fact that it's through the backup servers instead of the iMessage servers makes no difference.
What's more, it's possible to do better without sacrificing usability. For several years Android has been end-to-end encrypting backups using the user's lock screen passcode, with protection against brute force attacks provided by hardware secure elements. https://security.googleblog.com/2018/10/google-and-android-h...
> The fact that it's through the backup servers instead of the iMessage servers makes no difference.
It makes a big difference. If I print out the texts I receive, it doesn't change whether the texting program is end-to-end encrypted. The same goes for backups. An unencrypted system-level backup doesn't mean that the program being backed up is failing at security.
It's bad that Apple doesn't let you encrypt your backups properly, but it's a separate issue.
What if the texting program has a built in feature to print the texts you receive and mail a copy to the company that wrote the program, and it nags you to enable this feature all the time, and most of your friends have it enabled? Because that's a lot closer to the scenario here.
> An unencrypted system-level backup doesn't mean that the program being backed up is failing at security.
iOS programs can choose how their data is backed up. iMessage isn't just getting its data stolen by iCloud accidentally. These backups are a feature of iMessage as much as iCloud. And besides, iCloud is made by the same company, it's not a separate entity.
> Would you say that no desktop app that saves its key can ever qualify as end-to-end encrypted?
I would say that no app can qualify as end-to-end encrypted if a large fraction of users send their data to the maker of the app in a form that can be decrypted by the maker of the app, regardless of the reason.
Turning off iCloud backup is not a genuine choice, because it means you lose everything if you lose or break your phone (there is no other way to back up your phone except iCloud backup, Apple does not allow third-party phone backup services).
There’s a good HN thread from earlier this year about that, but basically, you can disable iCloud Backup and enable Messages in the Cloud, so that all of the messages are still backed up and synced between your devices but the keys are not, so that Apple can not read them. Then you can back up to your Mac/PC instead.
But most end-to-end encrypted apps aren't configured by most of their users to send their messages and encryption keys directly to the author of the app. iMessage is.
Feel free to trust who you want but I don't think Apple should be able to get away with calling iMessage end-to-end encrypted when they have most iMessages stored on their servers and the keys to decrypt them.
> At some point you have to trust somebody, right?
It's possible to use an actual end to end encrypted app that doesn't have the keys to read your messages stored on their servers.
I think this article is a bit over my head, but if Apple never has possession of users' private keys, how are they able to recover iMessage conversations when a phone is lost/stolen (which I know they can do)?
Zoom is such a bizarre product. For huge video calls, it tends to perform as well or better than everything else out there. Yet at the same time it literally seems like straight-up malware and seems to violate your trust and privacy left and right every step of the way, even in the installer (!).
To be fair it does perform better than everything else, which is why people are so forgiving of it, but it still doesn't excuse their ineptitude on privacy and security.
In my experience Google's Hangout Meetings have been at least as good or better quality and the interface is far superior in my opinion. For example it works in the browser without any plugins (even in Firefox.)
I haven't used Hangouts on a professional setting in a while. Does it finally support tile view? Another feature I find really valuable is allowing two windows. One for participant view and a separate one for shared screen.
yeah we use it. Not sure if this is still hangouts or something else underneath, but meet.google.com is the url which allows you to quickly create a session and share it. You can present your desktop using a separate tile.
My company uses Google meets and it works flawlessly with more than 100 people in a call. Automatic integration with GSuite and Chrome box are an added bonus.
Does it have the advanced features present in Zoom? At my $WORKPLACE we use the Spotlight feature so to ensure everyone is watching the right person (fairly important as we have lots of Deaf people and they all need to be watching the person signing) and Breakouts to set up discussion groups. While we don't have 100+ participants, we do have up to 25 active participants with lots of turn-taking (and not e.g. 100 watching 1 stream). Other platforms we tried didn't have these and made it difficult for participants especially those who aren't great with tech. Also helps Zoom seems to work well with video—other platforms we tried were great for audio so fine if you just need to listen to people but woeful with video making it a no go for meetings involving people who use sign language.
It was one of the fastest-growing SAAS companies before coronavirus hit despite all the free competition.
The reason? "It just works."
It's clear their singular focus on making it "just work" for even the least tech-savvy users has led them to prioritise user experience over security/privacy. I imagine a rebalancing is coming.
The age old question: is it bad because there are bad actors involved or is it bad because it's new and cheap and the maker views security etc as a Nice to Have.
IDA Pro is an reverse engineering tool sold by Hex Rays, which, for the last decade or two has been the the go-to software for security research, with very few actual competitors in the area (though this has been changing recently, depending on who you ask and what your needs are). It is fairly infamous for its exorbitant price (many thousands of dollars) and the company's extreme willingness to go after "pirates" (a term which is somewhat loose and is quite a bit broader than "people who use IDA Pro without purchasing it") and their inclusion of fairly invasive DRM reminiscent of malware.
Let me just add (in addition to what you wrote, all of which I've experienced) that I literally (using the term correctly here) had to /beg/ them to take my money ($7500, I think).
Very long story, but long story short, I needed the tool urgently (this was pre-Ghidra), like hours & days mattered. I am a small company, they were "skeptical" of me and my intentions, they demanded all sorts of paperwork and IDs, etc. It was seriously as though it was a dark comedy nightmare.
Everything you wrote is true. So many people fellate IDA given the chance; I don't know what they're smoking. The product is clunky and ugly (but it mostly works), but as you said, it's got very little competition. I and a few of my colleagues (and many here on HN) could probably do a better job, but I don't want to live in that world.
Man, I think I got pretty far afield there, but your comment resonated with me. Thank you.
Oh yes, I have heard this too. I know some people who have been “blacklisted” from ever buying IDA because Hex Rays thought they had “associated” with people with pirated copies. They are very selective with who gets the software and how they can use it…
The problem with Zoom isn't, that it doesn't have end-to-end encryption. If you consider the problem, it is extremely difficult to solve, as in contrast to a 1:1 video call, the server somehow has to to multiplex the streams and thus at least some access to them. With a trustworthy provider, this isn't a big issue. A trustworthy provider will have a clear policy which ensures that only the multiplexer has access to the unencrypted data. After all, you are running a proprietary client from that provider, and each client of course has access to all unencrypted data.
The problem is, that the actions of Zoom doesn't make them look like a trustworthy provider. They lied about the end-to-end encryption. What they should have done instead is to be transparent on how unencrypted data is used on their servers and what their protocols are to prevent unauthorized access to that data. Which is especially important in a business context, because the business users themselves have confidentiality agreements, they need to guarantee and using an external provider for confidential data required that provider passing the neccessary scrutiny.
And of course, the huge pile of security issues coming up with their client, the web server, the mac installer, the script host, give any reason to believe that they either don't know what they are doing or completely reckless at least. And the term "reckless" doesn't fit in a conversation about security :).
No. "End-to-end encryption" does not protect metadata necessary to route the data over a network, just the contents of the communication. The clients could negotiate keys to protect the contents of a meeting end-to-end. In other words, Zoom servers could deduce who was speaking, when they spoke, and for how long, but not what they said.
Your internet service provider can deduce the same about your HTTPS connections.
> The clients could negotiate keys to protect the contents of a meeting end-to-end.
Not really because Zoom makes fairly extensive use of the decrypted video streams on their servers, e.g. to detect who is talking, pause video for people with slow connections, etc. You could maybe do it for meetings with a few people in, but good luck doing it for meetings with 100 people.
Hell the cryptography of group end-to-end encryption hasn't really been worked out yet. WhatsApp doesn't do it and that's just for text. I'm pretty sure Signal doesn't either.
There's really nothing bad with not having end-to-end encryption for group video conferencing apps. The shitty thing is that they pretended that they did.
> Not really because Zoom makes fairly extensive use of the decrypted video streams on their servers, e.g. to detect who is talking, pause video for people with slow connections, etc. You could maybe do it for meetings with a few people in, but good luck doing it for meetings with 100 people.
You encrypt audio and video streams separately. If your connection is slow you stop grabbing the video stream. Detecting who is talking is a local function because you are receiving audio packets from them. Encryption doesn't have to change the amount of data sent.
> Hell the cryptography of group end-to-end encryption hasn't really been worked out yet. WhatsApp doesn't do it and that's just for text. I'm pretty sure Signal doesn't either.
Two ways:
1. Your password does not need to be shared with Zoom for entry into the meeting room. That password and the meeting room number are converted into a symmetric encryption key, so everyone who is able to join the meeting successfully has the same group shared secret for both sending and receiving video and audio.
2. You have an invitation system where the host of the meeting approves people. This approval causes the symmetric key of the meeting to be shared with the person seeking approval, encrypted to their public key. People who can supply a proof of possession of the password or whose public keys were associated with the meeting get invited without requiring confirmation by the host, although the participants in the room becomes a log of who was able to view the meeting.
These get harder if you want to say have a periodic key rotation while a meeting is going, for sure. They also get harder if you try to encrypt routing metadata or disguise that the traffic is audio/video in general.
The real thing that kills E2E for a corporate product like Zoom are the phone dial-ins. Hardly a point for all that security if you have one person calling in having the now unencrypted voice traffic bounced all over the place.
> There's really nothing bad with not having end-to-end encryption for group video conferencing apps. The shitty thing is that they pretended that they did.
100% agreed within Zoom's market because of integration of external services like dial-in/dial-out voice.
I presume they also resize video, so that you download small streams for thumbnails, plus a bigger one for the main view. If it's just two sizes you could do it before uploading, but this would be harder if it's actually 10 different qualities.
Zoom also has a feature that lets people join a meeting by phone. There's no way they could do that without being able to access the unencrypted stream.
The server does not have to decrypt or process the video, it can just send the same encrypted video to other clients (or you can use p2p communication without a server involved).
There may be advantages to processing video at the server, but it's definitely not a hard requirement.
The clients could control which video streams they want to subscribe to without the server decrypting them.
If you want to support thumbnail/fullscreen versions of streams, the clients could just send along two streams or use a codec that supports this kind, like h.264 SVC.
Yes, with effective 2 streams per client, the client could tell the server to send max. 1 full resolution stream and any number of low res streams. If you want full encryption, they have to be send as separate streams from each client. Also, the encryption would need to be negotiated between each pair of participants separately, that is doable, but scales with the square of all participants. So yes, full end-to-end encryption is possible, but quite an effort.
But as I wrote before, just having encrypted connections to the server should be fine, if the server provider is trustworthy.
Now we've moved into the territory of what's easier to implement, from what's possible. That's a different discussion and involves also the lack of implementation complexity at the server end.
There is no problem doing key exchanges between all the participants as long as the number of participants isn't too huge. Everyone just needs to kex with everyone else, so from the POV of a single participant the effort scales linearly with number of participants, even though the work collectively done by everyone is squared. It's the same setup as host-to-host ipsec with keying based on common CA between the peers.
The problem is the required bandwidth, if you send all streams to all participients. If you do that, you can do end-to-end encryption, but you have to send all streams at the maximum resolution you want to support. If the server handles the mixing, each client only needs to be send one full-resolution stream.
Matrix/Riot, Jitsi, and Signal seem to have handled E2EE voice/video pretty well while being FOSS (the first of which is a free/open protocol as well as a client).
That's the right thing to do. If SpaceX is important to national security, then Zoom security and privacy is so bad, that a bad actor could steal SpaceX technology.
At my previous job, we used to dial in random zoom numbers and entered into random conversations of other companies. Once we landed into a Facebook call where they were talking about Libra (before it was a thing).
If you turn of camera and video, the host doesn't even know you're there unless they check guest list.
I looked into adding Zoom to our Slack workspace this morning, and was beside myself with the set of permissions they requested — reading the contents of every channel and private chat they're included in? For a slash command?
That's a hard no. Turned me off the service entirely.
> reading the contents of every channel and private chat they're included in? For a slash command?
This may improve soon. Slack is starting to force apps to request granular permissions (vs a big-tent "bot" scope like before) and when you submit to their store, they vet each permission and verify what you're using it for. They don't let you request permissions "just because" in my experience.
It works, but it burns through so much CPU your computer will be a gibbering mess. There's some pretty silly inefficiencies going on, for example; if you switch away to another window, they display a small video player while keeping the big one running the background. Each time you switch into screen sharing mode, they drag you back to the app again. If you draw on the screen, someone has screwed up their linear algebra so you end up seeing double with an extra copy of what you're drawing, in totally the wrong place.
Annoyingly it's a bit too convenient, so going out of band is a pain.
I regularly have complaints that people can't hear me over my laptop's fan when using slack video conferencing, or jokes about hair driers. Computer is a MBP16.
Seems like they need to get their CPU use under control.
You could also get a headset instead of yelling at your built-in mic. As a 100% remote employee, even before Covid, the coworkers with headsets are much easier to understand, and have less background/foreground noise.
As a headset user myself, no one ever asks me to repeat myself, notices when I type, or hears anything not within an inch of my mouth. Plus, I've got a physical mute switch for instant, unambiguous mute, as-needed.
It works great for 1:1 and small group conversations, but its max meeting size (15 people I think) is an immediate disqualifier. We regularly have hundreds of people on Zoom calls with no issues.
YMMV, but for me ut turns my computer into a simulation of a space heater even when I'm not sharing my screen, and those gigantic unhidable toolbars makes me feel like I'm playing whack a mole more than working a times.
Oh, and yes,I "almost forgot" ! How to they even physically manage to get multiple seconds latency at times, when everyone in the call is in the same city, and where the worst connection on the call is at about 20Mbit? Are they just buffering for the heck of it, why not simply drop some frames if you get behind?
We have Zoom on our Slack workspace and we'd remove it immediately if this were the case, but it appears to be false. The full list of permissions required by the official Zoom Slack integration is at www.slack.com/apps/A5GE9BMQC-zoom, and doesn't have read access to any channels, private or public, except for "some URLs in messages".
I feel very conflicted on Zoom. It mostly just works and on every platform, even Linux. My kids use it to do their music lessons now. It really is very good at what it does in a time when such solutions are needed. Also they responded to the mobile facebook sdk issue and the macos issue quickly.
But I agree the way they suggest it is end to end encrypted is misleading. I don't think it really can be end to end to get the performance and features. People just need to see each other at the moment. You can do anything sensitive with more secure communication. But it clearly doesn't belong in any place discussing technology with military applications.
I still think it is solid for my kids to keep up with their lessons or for a weekly meeting about some web development. There are genuine criticisms of Zoom at the moment that need to be taken seriously but there is likely also some negative media being generated from their competition that are missing out.
They did not address the OSX issue for ages. They claimed it was intentional and a valuable feature for customers. In the end Apple had to release a security update to remove the web server.
So what's a good open source replacement? Jitsi, Jitsi Meet, Linphone, Ekiga, Jami and a bunch of others look okay [1], but it's hard to say how easy they are to use.
I've used Jitsi Meet with ordinary (that is, non-IT) people without trouble.I've also recommended it to elementary school teachers, who have been happy with it. It works with desktop browser, on mobile it offers a minimal app installation. Self-hosting is quite easy with their apt repo.
Came to say the same thing. Zoom is impossible if you are considered part of the DoD supply chain. Most of the public products are untenable if you need to talk about ITAR projects or about anything considered CUI.
Even if they weren't supplying the DoD with anything, rocket technology is missile technology. (Granted, liquid fueled ICBMs are old-school, but they'd nevertheless still be under ITAR even if they took no government contracts.)
I'm surprised by the concerns raised by the HN commenters. When Zoom filed their S1, HN was nothing but complimentary about their product and business model.
The time to pull out might be after their next quarterly report.
Spacex is a targeted environment. It's smart for them to not use such a service. Many years back it was getting hardware shipments intercepted and bugged. A company was spun out of just dealing with the amount of attempts to root Elons devices.
"In house (software)" is usually synonyms with "not properly tested or secured"... I'd generally rather they relied on third party code audited by the nations security services.
Its not so much that you need to have your own, its just much more complicated to do it in the cloud. When you have to be able to prove by audit that only US persons are authorized to access the datacenter and that only US persons have accessed the datacenter then you need a provider willing to do that as well as contracts, policies, and procedures. It all needs to be audit-able and evidence based. If you need to do these things in the cloud then you need to do them in-house too. Which means that you already do the basics in-house making it attractive to do things in-house.
Take Confluence for example, if I use the cloud version I can't store any files there or have any information about certain projects. But I can run the same Confluence in-house and then we can share project details. As you point out, its not always that easy.
Zoom's hidden doctrine "convenience over privacy" works really well. They assume their user wouldn't care too much about security and privacy, and sadly, for the vast majority, they assumed correct.
You don't get zoom's government platform by default, you have to ask for it and I'll bet it has different pricing. I suspect you also lose some features, like recording (at least at the server). I haven't checked Zoom but other FedRamp products I've seen are 3x the price of the standard offering.
FedRAMP moderate doesn't require an isolated, US citizen-only-staffed, environment. Only FedrAMP High requires that, so they can still be on shared infra for moderate.
FedRAMP Moderate is insufficient for ITAR work, that's true. The problem is that in order to get the appropriate rating, one of the things you have to do is go through the FedRAMP process, and only Slack, Zoom, and MS Teams are even trying.
Mattermost is self-hostable, so as long as you tightly control access, a lot of the SaaS prohibitions don't apply.
It was clear from Zoom's security vulnerability last year that they value ease-of-use over security, so it seems obvious to me they wouldn't care about privacy.
Setting up Jitsi on a very basic Linux VM is super easy. I know, because I am terrible at server setup and even I was able to do it in a couple of hours this morning.
I've found Hangouts to be equal in call quality to Zoom, and it runs entirely in the web browser. The only down side is it consumes a LOT of power, so you need to stay plugged in while you use it.
Consuming a lot of power could be a show stopper for organisations where lots of staff are on older equipment? Zoom works flawlessly even on older hardware.
You can have 200, and you can also live-stream large meetings. It is a bit weird that you have to split the stream, but it's worked great for us doing ~300 person all-hands meetings.
I'm not Zooming, but everyone I know is, and it looks pretty damn slick and seems to work great. I can't think of any service that has had as fast and huge an adoption, which is obviously due to blind luck/things beyond their control (no one saw global apocalypse coming and hoping for such would be a horrible business plan).
Of course there's been other services working just as well or better for at least 10 years now. Interesting things converged suddenly on Zoom. I guess full cross-platform support was the key? Also interesting is the deluge of anti-zoom articles that are blanketing the cybersphere :-) just as they are getting traction. Jealous rivals? Disgruntled lovers?
Zoom is the deluge du jour (literally—today has been the day of Zoom deluge complaints), but Occam called and wanted to let us know that we don't need any sinister explanations. It's obviously (or at least, explicable) as an effect of the covid crisis.
Has anyone noticed how hard some late-night-US-TV-shows are pushing (advertising) Zoom?
Stephen Colber, Seth Meyers...
Zoom is spending big $$$$ to advertise. While competition (WebEx for example) has been silently killing it. I feel that Zoom wants to be "the new cool kid" in town for which nobody knows anything about.. and while we do learn it looks that it's got some shady practices (as we discuss here).
I use MS Teams in my company, seems to work very well, great quality video, I’m don’t fully understand the security issues with Zoom, so I’m curious, does anyone know how MS Teams compares to Zoom??
The school my kids go to (in Portugal) also use Teams, for online learning, seems to be working well for them now. (Europe had issues a few weeks ago with Teams, when most of the schools first closed)
The issue with Teams is the issue all modern Microsoft services have: Their account systems are really confusing, and they actively make it worse in the UI.
I have 4 different accounts that have access to different Teams instances. They all have my full name, and therefore the same initials. There is no way to tell which account you are logged in as, in the UI.
A couple of months ago, they finally fixed the "log out and log in as a different user" flow so it actually did something for me. Before that, it would log out, and then just automatically log back in as the same user.
You can't (as far as I can tell) run two instances as different users at the same time.
I join most Teams meetings as "guest" on my phone. That works OK, but it doesn't really say much good about the Windows app.
MS Teams is better but you have to pay for it (and maybe it's complex to implement too).
Zoom is free (and supports large numbers of participants which other free alternatives don't). Also meetings can be setup with about 4 mouse clicks by the user himself with no other kit than a browser in Zoom. Getting your art department home workers to correctly
It has been weird watching people insist we use zoom because it supports so many users, when there are only three of us actually in a meeting. So the cynic in my suspects it's being used by (say) schools and universities (as the Skype limit was 32 people, now 50). And other people have assumed it's better since Harvard\Cambridge use it, so now they use it for 4 people...
Thanks - Teams is surprisingly easy (for a Microsoft product) to set up a glitch free video conference call.
I was curious how secure it is?
Its free with an Office 365 subscription and they are quite cheap, office is pretty much essential for most businesses, I think schools get it very cheap or even for free.
I've never heard of the Zoom app until about three weeks ago and now I'm reading something about it here on HN almost every day. Why is this company and their app suddenly so popular and so critiqued? I thought the world was Skype, FaceTime, WebEx, that whole cluster of technologies and related applications. Why are the Zoom folks suddenly on the front page everywher?
WebEx is terrible, FaceTime is Apple-only, Skype is circling the drain since being bought by Microsoft, so there is a lot of pent-up demand for a good cross-platform videoconferencing tool. Zoom is one of the front-runners.
There is a good overview of the options in an article recently posted here - see under "Videoconferencing":
Being the Linux geek, I use MS Teams. I even use it with its Linux client, and works astonishingly well.
My only 2 grievances with it are:
1. Teams steals focus to make the next message on a group, rather than in the threat ALL THE TIME. Ive been there for a dozen comments because Teams stole cursor focus.
2. Its easy to make an invite to a one-shot room, rather than use an existing room. Doing so loses all history and provenance and discussion. And there's no "merge this room with the real place" when that happens.
But all in all, Ive been on 130+ person calls with no issue. Works very well, aside those issues above.
Teams has video conferencing? Wow never knew this and I use Teams often for $WORKPLACE. I find it a bloated mess with little things that don't work very well e.g. opening Word documents opens in a weird HTML editor which stuffs up the formatting of delicate documents. It's so bad I never share things via Teams and just email instead. So not sure if I'd want to try its video conferencing function if it can't get other basic stuff right?
Remember the movie Antitrust, where Bill Gates type character played by Tim Robbins records everyones computer screens to steal secrets? Thats Zoom for you, minus the murders.
And the amount of classified material they handle, and the fact that Jeff Bezos is a direct competitor so Amazon is very very closely affiliated with a direct competitor.
This is a knee-jerk reaction. Zoom was indirectly sending information to Facebook as part of their Facebook SDK integration to enable login, not directly as part of explicit information gathering. Any app that integrates with the Facebook SDK does this, and they've also patched it already [1].
The supposed root exploit found in Zoom also requires physical, logged-in access to the machine, at which point a Zoom exploit is the least of your problems [2].
Zoom is a solid piece of software, and the developers are responsive and seem to care. I'm disappointed to see it getting dumped on during the past few days. A cynic might even suspect a co-ordinated campaign by Cisco, considering Zoom was started by frustrated ex-Cisco employees and has had runaway growth during the viral crisis, while the same cannot be said for Cisco's competitor product WebEx.
> Zoom is a solid piece of software, and the developers are responsive and seem to care.
I get the exact opposite impression. Not just due to these bugs, but also the hidden webserver thing [1] a while back.
Recurring theme in this (the webserver and the installer issue) seems to be an unhealthy obsession with reducing the number of clicks the user needs to perform. They deliberately chose dubious, hacky solutions over doing things the right way due to this. It makes you wonder what other bad decisions they made in the client or server code.
> Zoom is a solid piece of software, and the developers are responsive and seem to care.
The dodgy things they've been doing suggest otherwise.
* Hijacking package preflight script rather than standard package installation mechanism, so their software is installed before the user clicks Install.
* Installing a hidden web server without user consent.
When you make weapons technology (as SpaceX does; rockets are weapons technology) and are involved in launching military satellites (as SpaceX is), you kinda have to take security issues seriously.
exactly, plus when you're spaceX or a similar tech company it's much better to be safe than sorry. you don't always have the luxury to assume good intentions.
Decisions like this have less to do with "what are you doing now" and "what was your intention", and more to do with "is there a pattern" which is an externally surfaced signal that something deeper is going on with their engineering organization, long term business plans etc.
Zoom has a pattern of abusing users' trust. Ok, maybe its not intentional, but increasingly, that doesn't matter in security discussions. The fact that it was unintentional is a concern itself. Maybe they lack discipline, or leadership, or something; I'm not asserting to know. But, its no longer good enough to simply say "we're sorry, it was an accident".
From the article it seems like Elon is mostly worried about meeting IDs leaking and letting any random connect:
> The FBI’s Boston office on Monday issued a warning about Zoom, telling users not to make meetings on the site public or share links widely after it received two reports of unidentified individuals invading school sessions, a phenomenon known as “zoombombing.”
This is certainly not a 'knee-jerk reaction' to the use of the Facebook SDK, but presumably extrapolates from the history of false uninstallation, insecurity by obscurity, false advertising, etc.
Each of those things has been, as you intimate, patched or resolved -- but does that give you increasing or decreasing confidence in their product and priorities?
> The supposed root exploit found in Zoom also requires physical, logged-in access to the machine, at which point a Zoom exploit is the least of your problems
No, it does not; 9to5Mac is just playing a game of telephone from the original blog post and the actual issue only requires code execution on the machine.
The rumor that I've heard is that the "security and privacy concerns" are that Boeing has somehow gotten access to their Zoom. I don't know what that means, exactly, but it's at least somewhat more rational than being concerned that facebook is siphoning up the data.
