4 ms·

Last time I looked Telegram wasn't recommended by Moxie and a few other people. That was 2-3 years ago. What's the status now? They came up with their own encr

by rdtsc 9y ago

Last time I looked Telegram wasn't recommended by Moxie and a few other people. That was 2-3 years ago. What's the status now? They came up with their own encryption protocol and they are not trained or known as cryptographers, that's a warning sign. > [Durov] The encryption of Signal (=WhatsApp, FB) was funded by the US Government. I predict a backdoor will be found there within 5 years from now... That's another red flag, needing to spread fud and lies about Signal. Another reason not to trust Telegram. Indicates that maybe their ethics and integrity are a bit too flexible. > During our team's 1-week visit to the US last year we had two attempts to bribe our devs by US agencies + pressure on me from the FBI. Some people might read it as "these guys are so good, FBI is begging to backdoor them". But it can also be read as FBI suspects they are ethically compromised and they have a chance of succeeding.

geofft 9y ago
> [Durov] The encryption of Signal (=WhatsApp, FB) was funded by the US Government. I predict a backdoor will be found there within 5 years from now... To emphasize even more why this is silly: nobody from the government (well, at least nobody who has said they're from the US or any other government) was involved with the actual development of the Signal Protocol. It was just funded by the government, through the Open Technology Fund, a project of Radio Free Asia (which is itself under the Department of State; the OTF was largely an initiative of Sec. Clinton). This is an extremely different part of the government from either NIST or any of the three-letter agencies. There has been exactly one backdoor found in crypto relating to the US government (Dual_EC_DRBG), and it was in crypto developed by the NSA and basically pushed into a standard. It was also crypto that looked extremely suspicious, immediately, to any cryptographer who looked at it: it had a contrived design for no good reason, ran much slower than the existing options in the space, and appeared to support a backdoor. A lot of people have looked at the Signal Protocol and found nothing like this. The US government has been accused of hiding backdoors before in one case, DES. It turned out that they were hiding a way of strengthening DES against an attack that was not yet public (differential cryptanalysis) but had been discovered by the NSA. Nothing like this happened to Signal: the developers were not told by the government "Great, just use these S-boxes instead." (In fact, it occurs to me that there's not much room in Signal for a backdoor along the above lines: no S-boxes, no constants, etc. The closest it gets to that is picking Curve25519 and Curve448, both of which are well-known curves, predating Signal, with simple mathematical descriptions that make them essentially impossible to have backdoors. Perhaps he means that the Signal software is backdoored? But that wouldn't make sense with the reference to WhatsApp and Facebook, and also is a much more easily disprovable assertion than that crypto is backdoored.) The US government also, of course, weakened crypto with the ridiculous export rules of the '90s. But that wasn't a backdoor, and they were pretty explicit that the intention was to weaken crypto. The OTF has no such motivation here; their goal was to produce secure tools that can be used by dissidents around the world, and intentionally-weakened crypto would be dangerous in such a context.
- abecedarius 9y ago
  I worked on three OTF-funded security audits a few years ago (though not of Signal). No spook crossed my radar at any time. I'd have been very surprised if any had tried to pressure us at that work; you can read Bamford or the Snowden stories to get a better idea of how they seem to work. (Despite having some security experience I'm not a cryptographer, and I feel kind of silly speaking up here. But the FUD is even sillier.)
Iv 9y ago
As someone who is not exactly versed in the jurisdictions of the TLAs, wouldn't it be the CIA's role to put backdoors in Telegram? The FBI would be doing counter-intelligence and would only approach Telegram if they suspected wrongdoings on their parts, no?
AFNobody 9y ago
> Last time I looked Telegram wasn't recommended by Moxie and a few other people. That was 2-3 years ago. What's the status now? I honestly don't think it matters. The fact they lie about a large number of items, such as the nature of their relationship/business and have repeatedly is proof enough they can't be trusted. https://www.washingtonpost.com/news/the-intersect/wp/2015/11/23/the-secret-american-origins-of-telegram-the-encrypted-messaging-app-favored-by-the-islamic-state/ https://www.washingtonpost.com/news/the-intersect/wp/2015/11... > For starters, Durov and other Telegram employees had repeatedly claimed their app was nonprofit, which wasn’t technically true. (“The Telegram team declared numerous times in its FAQs and public statements that Telegram was a non-profit,” Durov wrote to Neff in an email made public in court documents. “… The for-profit entity that we currently have, especially a U.S.-based one, raises questions among our audience.”) > That's another red flag, needing to spread fud and lies about Signal. Another reason not to trust Telegram. Indicates that maybe their ethics and integrity are a bit too flexible. Lying has been their modus operandi since the start. https://news.ycombinator.com/item?id=6948742 https://news.ycombinator.com/item?id=6948742 > You must not have followed Telegram much. From the beginning they've done nothing but pretend their protocol is absolutely secure ("military-grade encryption", "world's most secure protocol", etc) and rejected any attempt from the crypto community to help them fix problems before they endanger people. The sheer pervasiveness of their lies and the fact it ranges from the quality of their cryptographic skills to the nature of their business should have prevented its adoption. Unfortunately, lying at scale appears to work because most people simply dismiss it as "competitors" attacking their competition. :\
s17n 9y ago
Or it can be read as "this is a made up story; the guy is full of shit."
tptacek 9y ago
It's the same status as it's always been. Nothing fundamental has changed about their technology, which is materially inferior to that of Signal, WhatsApp, Wire, even Facebook Messenger.
travmatt 9y ago
> [Durov] The encryption of Signal (=WhatsApp, FB) was funded by the US Government. I predict a backdoor will be found there within 5 years from now... If you can't win an argument on the merits, spew unfounded sensational accusations to confuse and distract the technically illiterate and conspiracy minded.
tripzilch 9y ago
> they are not trained or known as cryptographers, that's a warning sign. Actually, they are. Or a few of Durov's family members are. IIRC it was his brother that developed Telegram's cryptography. His family has a pretty impressive academic mathematical background. The warning sign is that, academic credentials or not, they still rolled their own crypto, and didn't use common ( = "crypto community" vetted/approved) cryptographic primitives and techniques, but came up with new stuff that isn't vetted properly. And that is a warning sign. What is also problematic is that Telegram's encryption is not enabled by default. Even if the crypto turns out to be flawless, that is still a big problem. One other thing I wonder sometimes myself--though I'm not knowledgeable enough to lean either way--is about this crypto community. I know it's international, in principle. But how inclusive is it in practice? I'm probably biased from reading most on the topic via HN, but how much of it is US-based? Does Russian academia take part in it too? Honest question. BTW I use Telegram (also Whatsapp, Signal and recently Matrix) myself. I just don't consider any of it (Telegram) encrypted. Really fast postcards. But the UI and functionality (especially with the desktop app) is a joy. Why can't they all just team up or something and have the Telegram guys do the UX part? :)

by rdtsc 9y ago

geofft 9y ago

> [Durov] The encryption of Signal (=WhatsApp, FB) was funded by the US Government. I predict a backdoor will be found there within 5 years from now... To emphasize even more why this is silly: nobody from the government (well, at least nobody who has said they're from the US or any other government) was involved with the actual development of the Signal Protocol. It was just funded by the government, through the Open Technology Fund, a project of Radio Free Asia (which is itself under the Department of State; the OTF was largely an initiative of Sec. Clinton). This is an extremely different part of the government from either NIST or any of the three-letter agencies. There has been exactly one backdoor found in crypto relating to the US government (Dual_EC_DRBG), and it was in crypto developed by the NSA and basically pushed into a standard. It was also crypto that looked extremely suspicious, immediately, to any cryptographer who looked at it: it had a contrived design for no good reason, ran much slower than the existing options in the space, and appeared to support a backdoor. A lot of people have looked at the Signal Protocol and found nothing like this. The US government has been accused of hiding backdoors before in one case, DES. It turned out that they were hiding a way of strengthening DES against an attack that was not yet public (differential cryptanalysis) but had been discovered by the NSA. Nothing like this happened to Signal: the developers were not told by the government "Great, just use these S-boxes instead." (In fact, it occurs to me that there's not much room in Signal for a backdoor along the above lines: no S-boxes, no constants, etc. The closest it gets to that is picking Curve25519 and Curve448, both of which are well-known curves, predating Signal, with simple mathematical descriptions that make them essentially impossible to have backdoors. Perhaps he means that the Signal software is backdoored? But that wouldn't make sense with the reference to WhatsApp and Facebook, and also is a much more easily disprovable assertion than that crypto is backdoored.) The US government also, of course, weakened crypto with the ridiculous export rules of the '90s. But that wasn't a backdoor, and they were pretty explicit that the intention was to weaken crypto. The OTF has no such motivation here; their goal was to produce secure tools that can be used by dissidents around the world, and intentionally-weakened crypto would be dangerous in such a context.

abecedarius 9y ago
I worked on three OTF-funded security audits a few years ago (though not of Signal). No spook crossed my radar at any time. I'd have been very surprised if any had tried to pressure us at that work; you can read Bamford or the Snowden stories to get a better idea of how they seem to work. (Despite having some security experience I'm not a cryptographer, and I feel kind of silly speaking up here. But the FUD is even sillier.)

Iv 9y ago

As someone who is not exactly versed in the jurisdictions of the TLAs, wouldn't it be the CIA's role to put backdoors in Telegram? The FBI would be doing counter-intelligence and would only approach Telegram if they suspected wrongdoings on their parts, no?

AFNobody 9y ago

> Last time I looked Telegram wasn't recommended by Moxie and a few other people. That was 2-3 years ago. What's the status now? I honestly don't think it matters. The fact they lie about a large number of items, such as the nature of their relationship/business and have repeatedly is proof enough they can't be trusted. https://www.washingtonpost.com/news/the-intersect/wp/2015/11/23/the-secret-american-origins-of-telegram-the-encrypted-messaging-app-favored-by-the-islamic-state/ https://www.washingtonpost.com/news/the-intersect/wp/2015/11... > For starters, Durov and other Telegram employees had repeatedly claimed their app was nonprofit, which wasn’t technically true. (“The Telegram team declared numerous times in its FAQs and public statements that Telegram was a non-profit,” Durov wrote to Neff in an email made public in court documents. “… The for-profit entity that we currently have, especially a U.S.-based one, raises questions among our audience.”) > That's another red flag, needing to spread fud and lies about Signal. Another reason not to trust Telegram. Indicates that maybe their ethics and integrity are a bit too flexible. Lying has been their modus operandi since the start. https://news.ycombinator.com/item?id=6948742 https://news.ycombinator.com/item?id=6948742 > You must not have followed Telegram much. From the beginning they've done nothing but pretend their protocol is absolutely secure ("military-grade encryption", "world's most secure protocol", etc) and rejected any attempt from the crypto community to help them fix problems before they endanger people. The sheer pervasiveness of their lies and the fact it ranges from the quality of their cryptographic skills to the nature of their business should have prevented its adoption. Unfortunately, lying at scale appears to work because most people simply dismiss it as "competitors" attacking their competition. :\

s17n 9y ago

Or it can be read as "this is a made up story; the guy is full of shit."

tptacek 9y ago

It's the same status as it's always been. Nothing fundamental has changed about their technology, which is materially inferior to that of Signal, WhatsApp, Wire, even Facebook Messenger.

travmatt 9y ago

> [Durov] The encryption of Signal (=WhatsApp, FB) was funded by the US Government. I predict a backdoor will be found there within 5 years from now... If you can't win an argument on the merits, spew unfounded sensational accusations to confuse and distract the technically illiterate and conspiracy minded.

tripzilch 9y ago

> they are not trained or known as cryptographers, that's a warning sign. Actually, they are. Or a few of Durov's family members are. IIRC it was his brother that developed Telegram's cryptography. His family has a pretty impressive academic mathematical background. The warning sign is that, academic credentials or not, they still rolled their own crypto, and didn't use common ( = "crypto community" vetted/approved) cryptographic primitives and techniques, but came up with new stuff that isn't vetted properly. And that is a warning sign. What is also problematic is that Telegram's encryption is not enabled by default. Even if the crypto turns out to be flawless, that is still a big problem. One other thing I wonder sometimes myself--though I'm not knowledgeable enough to lean either way--is about this crypto community. I know it's international, in principle. But how inclusive is it in practice? I'm probably biased from reading most on the topic via HN, but how much of it is US-based? Does Russian academia take part in it too? Honest question. BTW I use Telegram (also Whatsapp, Signal and recently Matrix) myself. I just don't consider any of it (Telegram) encrypted. Really fast postcards. But the UI and functionality (especially with the desktop app) is a joy. Why can't they all just team up or something and have the Telegram guys do the UX part? :)