Closed dav-m85 closed 3 weeks ago
virtadpt
commented
2 years ago The last time I paid attention to Telegram was when Jakob Jakobsen did their master's thesis on breaking Telegram's encryption.
I don't know if anything in this paper has been remediated, but I'm inclined to say "no."
Lissy93
commented
2 years ago So I've not checked up on this in a while, so I'll re-look into it and update you here and the content in the repo :)
But the short story is that the backend is not open source (not a deal breaker alone), but also encryption is not on but default (you must enable Secret Chats), and this isn't supported for group chats. Their MTProto encryption has also been criticised over the years, but this might have improved now.
It's also had a bit of controversy in the past regarding logging, censorship, the company itself and the published app not being the same version as on their GitHub.
At the end of the day, it depends what your threat model is, and what your using it for. It's really hard to find a great all-round secure messenger, and to then get all of your contacts on-board. Signal is a good balance between ease of use and security for most, but again far from perfect.
Lissy93
commented
2 years ago Heya @dav-m85, I've done a bit of research, and believe that Telegram should remain under the Word of Warning section.
That's not to say it doesn't have it's place, or that it's actively dangerous to use, I don't believe either of those are true (Telegram is great for a lot of things). The biggest issue seems to be that it's marketed as a private and secure messenger, which can lead to a false sense of security. Below is a summary of things I looked at to come to this decision:
After re-reading Jakob Jakobsen's article On the CCA (in)security of MTProto, it looks like the first attack described (Padding-Length Extension) has been patched, but I could not find any info about a fix for the second issue (Last Block Substitution). That said, they're rolling their own crypto (MTProto), it's had issues in the past, and will new issues will likely be found in the future.
I looked into the apps too. The latest version of the Telegram Android app (appears to) have no trackers, and this is verified also on Exodus - which is great. It does however include 50+ permissions, but you have control over which of those you grant. It's awesome that both their Android (source) and iOS (source) apps are open source (but cannot be certain what you're running is the same as in the repo)
Then I read through their updated Privacy Policy, and a couple of things jumped out at me...
Section 5.2 stats that "we may collect metadata such as your IP address, devices and Telegram apps you've used, history of username changes, etc.". This is in the name of spam-prevention, but the data is still collected and stored.
Then sec 5.3 it states that Telegram's "moderators may check messages that were reported to them"
Sec 5.4 states "We may also store some aggregated metadata to create Telegram features", and this is followed up in section 5.5 where they say they collect "data about how you use Telegram" to influence certain features, but it's a bit vague. They don't mention if any of this metadata is encrypted, so I guess that it is stored in plaintext.
Sec 8.2 states "We may share your personal data with: (1) our parent company, Telegram Group Inc, located in the British Virgin Islands; and (2) Telegram FZ-LLC, a group member located in Dubai, to help provide, improve and support our Services" (but I'm not too familiar with these jurisdiction's privacy laws)
Sec 8.3 states that Telegram "may disclose your IP address and phone number to the relevant authorities." This is (I guess) unlikley to happen in practice, but the fact that this data is collected means it could be exploited.
sec 7.1 states that Telegram "relies on different payment providers around the world", so if you ever ineract with a bot that has the payment-platform enabled, then that's a whole new privacy policy to check.
sec 4.1 mentions that data is stored in "third-party provided data centers in which Telegram rents", so a physical attack isn't out of the question.
Sec 4.3 says "the personal data that you provide us will only be stored for as long as it is necessary", but then section 10 which is regarding deleting data seems very good (but again no way to verify).
Sec 6.3 about bots, and states what data third-party developers can get access to if enabled, they have access to messages in group chats, and can get your IP if you click a link. Both probably obvious.
Telegram have talked about the lack of on-by-default end-to-end encryption in quite a few of their blog posts. But their approach seems to be to dodge the topic, and the way they describe it is quite deceptive. They also seem to focus more on kicking down other messengers and disregarding reputable research, with very little hard evidence to backup what they're saying. Take this post as an example
As mentioned previously, none of this is to say that you shouldn't use it. The aim of this repo is to present the facts, and let users decide what's best for them. It's really hard to just say that an app is either good or bad for privacy, because there's so many other factors, in this case what you're using it for, and what your threat model is. And with Telegram, tbh unless maintaining water-tight privacy & security is something that your life or livelihood depends on, then chances are you will be totally fine using it. And it's still a hell of a lot better than WhatsApp ;)
If anyone else has anything else to add, or if I've missed anything/ got anything wrong, then please do let me know below :)
SadMadLad
commented
3 weeks ago Closing this for now. Please feel free to open it if there are any follow up queries.
Lissy93 wrote 2 years ago that Telegram was deemed untrustworthy since its source code is not public, here.
As of today, it's not the case anymore, see https://github.com/TelegramMessenger
Is there any reason to maintain Telegram (and perhaps the other warny, I haven't checked) in this caution section? If yes, an elaboration would be adequate IMHO (and in fact I'd like to know since I'm using it, if I should stop or not).
Thanks!