9 ms·

This bill is great in that email stored on your ISP's (or Google's) servers is treated identically to that on your computer at home-- they need to get a warrant

This bill is great in that email stored on your ISP's (or Google's) servers is treated identically to that on your computer at home-- they need to get a warrant. That's a great step forward. It isn't BS. However, it doesn't address intercepting and recording communications in-transit, which is extremely commonplace these days. Technically american intelligence agencies need a warrant to eavesdrop on american citizens. But in practice, they can get around this by recording _everything_ (full take), indexing it, and limiting their searches against communications by US citizens where they either talk to a foreigner or match a search query on a nefarious keyword. [1] They wave away violating the 4th amendment of the US constitution by saying that copying, storing, indexing, and searching through US citizens' private information is OK as long as a human doesn't read it. "Full take" sounds incredible-- it entails recording and retaining EVERYTHING passing through the internet. But we know it's possible. Snowden revealed that the UK GCHQ has been doing it since at least 2013 with their Tempora system [2]. Snowden also revealed the NSA's very similar XKeyscore system. At the time of his leaks, the NSA did not do Full Take... but that was several years ago. It's very reasonable to assume that everything, EVERYTHING, you do on the internet is recorded and stored by the government. This bill does not address that. That's why it's absolutely critical that everything you do is end-to-end encrypted. These government agencies have the resources to crack most encryption, too [3] but at least you aren't making it trivial. [1] https://en.wikipedia.org/wiki/XKeyscore https://en.wikipedia.org/wiki/XKeyscore [2] https://en.wikipedia.org/wiki/Tempora https://en.wikipedia.org/wiki/Tempora [3] https://en.wikipedia.org/wiki/Bullrun_(decryption_program) https://en.wikipedia.org/wiki/Bullrun_(decryption_program)

jcranmer 10y ago
> However, it doesn't address intercepting and recording communications in-transit, which is extremely commonplace these days. MTA<->MTA SMTP transmission is increasingly SSL these days, and end-user MUA<->MTA transmission has pretty much been entirely SSL for a long while. Google cites about 80-90% of its MTA<->MTA transmission is encrypted (https://www.google.com/transparencyreport/saferemail/ https://www.google.com/transparencyreport/saferemail/).
- hendersoon 10y ago
  Yes, see my link to the BULLRUN program. There were quite a few follow-ups as well. [1] https://freedom-to-tinker.com/2015/10/14/how-is-nsa-breaking-so-much-crypto/ https://freedom-to-tinker.com/2015/10/14/how-is-nsa-breaking... [2] https://arstechnica.com/security/2015/10/how-the-nsa-can-break-trillions-of-encrypted-web-and-vpn-connections/ https://arstechnica.com/security/2015/10/how-the-nsa-can-bre... [3] https://blog.cryptographyengineering.com/2013/12/03/how-does-nsa-break-ssl/ https://blog.cryptographyengineering.com/2013/12/03/how-does... And so on and so forth. Also keep in mind that they are retaining a full take online-- storing all traffic on the internet. Even if it can't be decrypted in realtime, they can come back to it in, say 10 years, with all the advancements in brute forcing and potentially quantum computing available in 2027, to take a look at your 2017 activity. The emails you wrote, the websites you visited, all your chats, all your searches, saved forever.
  - willstrafach 10y ago
    There is no evidence indicating that anyone can break encryption for mail servers using STARTTLS.
    - Karunamon 10y ago
      Not right now, no. But if you start saving encrypted streams, and then come back to them 10 years later when, perhaps, a TLS vulnerability is found...
      - willstrafach 10y ago
        I don't know what the media coverage stated, but the source material indicates that full-take is only stored in a 5 day rolling buffer. I believe 30-day for certain flagged selectors. Not at all indefinite, that is not possible.
        hendersoon 10y ago
        As of 2011, GCHQ stored their full-take for 3 days and metadata for 30 days. Do you believe they haven't increased capacity in the past 6 years? Again, the discussion comes down to evidence versus reasoned speculation. Evidence of intelligence agency activity is generally pretty tough to come by.
ocdtrekkie 10y ago
> They wave away violating the 4th amendment of the US constitution by saying that copying, storing, indexing, and searching through US citizens' private information is OK as long as a human doesn't read it. This is funny because this is the same technicality companies like Google use to claim they aren't violating your privacy: That the only thing reading your email is their data mining algorithms, not humans.
- srean 10y ago
  That's a strained analogy. Its far easier to not use Google. I have to actively agree to Google's terms and conditions. OTOH I have to actively opt out (if at all that is possible). Ideally in a democratic forum I would have some say in it, but that exists only as a fiction.
  - sokoloff 10y ago
    How many emails do you get from / send to gmail users? For me, it would be difficult to avoid using google because so many of my correspondents use it.
    - srean 10y ago
      I am saying that legally disowning your country, extricating yourself out of NSA coverage etc would be more difficult.
greydius 10y ago
I suppose it's a good time to invest in hard drive manufacturers.
Spooky23 10y ago
You need to separate national security activity from traditional law enforcement activity.
- givinguflac 10y ago
  Why? Parallel construction proves that law enforcement does not.
  - cptskippy 10y ago
    Does not what?
    - hendersoon 10y ago
      They don't separate national security from law enforcement. They _say_ they do, but they don't. Parallel construction is when an intelligence agency tips off law enforcement about some nefarious actor (like say, Ross Ulbricht) and then LE has to construct a plausible scenario where their hard-boiled detectives tracked down the bad guy through traditional police work, not through spying on American citizens. [1] https://www.washingtonpost.com/news/the-switch/wp/2013/08/05/the-nsa-is-giving-your-phone-records-to-the-dea-and-the-dea-is-covering-it-up/ https://www.washingtonpost.com/news/the-switch/wp/2013/08/05...
willstrafach 10y ago
> match a search query on a nefarious keyword. Sorry this is not true. You are referring to an email chain involving FI maybe, and within that USPI would be minimized so that the analyst cannot see who the US one is. They would need to specifically override and would need a good reason to get USPI as that is taken seriously. > "Full take" sounds incredible-- it entails recording and retaining EVERYTHING passing through the internet. This is also not true. It sounded nice during the leaks and everyone loved to hear it, but this is simply not true, not plausible, and would be a huge waste of money because they would wind up with too much USPI (totally useless). The collection systems are mostly going to be getting traffic from MENA area, although the leaked information was vague regarding specifics, for obvious reasons. (No argument with regards to ethics of bulk SIGINT collection, many opinions out there on that, just commenting as it is very unfortunate to see falsehoods repeated due to reporters poorly understanding the source material)
elbii 10y ago
> This bill is great in that email stored on your ISP's (or Google's) servers is treated identically to that on your computer at home-- they need to get a warrant. That's a great step forward. It isn't BS. Except for one major difference: Gag orders. When a warrant is served to Google to access your email, you may never know about it. If they show up at your home and take your email server, who you gonna gag? Unreasonable or overly-broad gag orders are one reason I always advocate running your own services whenever possible (granted, this needs to be easier).
- schoen 10y ago
  The distinction you mention is real, although not all warrants against providers come with gag orders, and some warrants against people's homes authorize surreptitious searches, such as by breaking into the home when the target is away. (This was done for a personal computer as early as 1999 in the Scarfo case.) Lately law enforcement has been seeking warrants to hack people's computers and some of them have been granted; typically those have also been surreptitious searches. I think this is the greatest risk for surreptitious government access to self-hosted information, because some agencies have been getting pretty excited about this and there are a lot of contractors who will supply them with vulnerabilities and tools even if they don't have the expertise to develop those themselves.
- hendersoon 10y ago
  That is certainly true. The difference is that before they didn't need a warrant at all if your cloud-stored data was over 180 days old. A court order alone was sufficient.
gnopgnip 10y ago
The legality of this is being litigated now. https://www.eff.org/cases/jewel https://www.eff.org/cases/jewel
loeg 10y ago
> It's very reasonable to assume that everything, EVERYTHING, you do on the internet is recorded and stored by the government. This bill does not address that. I mean, that's hard to believe. I work for a large storage vendor. If this was true, the feds would be our largest customer by far. They aren't. I don't think SE Asia makes enough disk to store everything indefinitely, and given exponential network traffic growth, even storing the most recent few years is a problem.
- dv_dt 10y ago
  I might consider that they're just not buying it from your company, or you aren't privy to the transactions to those gov't customers. How much storage does 1.5 billion buy? (Ok it's not all just storage...). https://en.wikipedia.org/wiki/Utah_Data_Center https://en.wikipedia.org/wiki/Utah_Data_Center The explosion of data you mention is a real problem; though thinking about it, one might manage with a mix of light filtering, leaning on organizations that already store the relevant data, etc.
- pbasista 10y ago
  That is just one opinion against another. With little fact to support any side. In your comment, for example, you seem to suppose that those agencies do not buy storage on secondary markets. Can you confirm that? Because if they do, then your sales statistics are almost meaningless in this regard. Also, it seems like you assume that everything the agencies collect is stored "as it was transferred" (i.e. if the same YouTube video was viewed millions of times, they would store it in millions of copies). I think it is safe to assume that these agencies are a little bit smarter than that and if they store everything, they have some kind of compression in place to minimize redundant storage of the same chunks of data. In any way, I think that until someone like Snowden speaks up again and brings evidence from inside of these agencies, we are only left with speculation.
  - hendersoon 10y ago
    There's no speculation on the UK. We know the GCHQ does this. Of course they generate quite a bit less traffic than the USA, given their smaller population. We don't know whether the NSA does "Full Take" today. That's speculation. I would be astonished if they didn't, and it is reasonable to behave as if they do. It's possible, it was possible 4 years ago, and storage is not getting _more_ expensive over time.
ThatGeoGuy 10y ago
> That's why it's absolutely critical that everything you do is end-to-end encrypted. These government agencies have the resources to crack most encryption, too [3] but at least you aren't making it trivial. This comment in particular seems to imply that government agencies can break e2e encryption. However, other sources would suggest that's not entirely the case. Snowden, for example, still seems to believe PGP works. I don't have reason to believe that it doesn't, or that the NSA can currently crack any arbitrary 4096-bit RSA keys. Further, there's a lot to be said about other modern encryption standards (ed25519 curves, forward secrecy and asymmetric double-ratchet protocols like Signal), which I don't believe have been shown to be compromised. I understand the importance of e2e encryption, but it seems that you're overstating the governments ability to crack it. I agree that SSL and TLS don't have the best history, and maybe we shouldn't rely on them, but it is also important to know that there are solutions out there which we still have quite a bit of reason to believe we can trust.
- jwtadvice 10y ago
  > This comment in particular seems to imply that government agencies can break e2e encryption. Much E2E encryption can be broken by government agencies. For HTTPS for instance, the government is a certificate authority accepted by your browser - as well as there having been multiple cases of collaboration with well known CAs to have certs minted for them. In this case there's no need to break RSA. In any case, much of TLS is open to various kinds of breakage (that is not as hard as RSA), and NSA docs showed a number of instances where intelligence agencies would sit behind load balancers and at data center chokepoints in domestic companies to gather information where PKI no longer poses a problem (behind the load balancers being referred to here as "end-to-end"). Basically, there's many ways to thwart "e2e" that isn't cracking 4096 bit RSA. Intelligence agencies do all of them.
  - hendersoon 10y ago
    Yes, very much this. Full circle back to the famous xkcd comic that's always posted in HN articles like this. https://xkcd.com/538/ https://xkcd.com/538/
  - 10y ago

jcranmer 10y ago

> However, it doesn't address intercepting and recording communications in-transit, which is extremely commonplace these days. MTA<->MTA SMTP transmission is increasingly SSL these days, and end-user MUA<->MTA transmission has pretty much been entirely SSL for a long while. Google cites about 80-90% of its MTA<->MTA transmission is encrypted (https://www.google.com/transparencyreport/saferemail/ https://www.google.com/transparencyreport/saferemail/).

hendersoon 10y ago
Yes, see my link to the BULLRUN program. There were quite a few follow-ups as well. [1] https://freedom-to-tinker.com/2015/10/14/how-is-nsa-breaking-so-much-crypto/ https://freedom-to-tinker.com/2015/10/14/how-is-nsa-breaking... [2] https://arstechnica.com/security/2015/10/how-the-nsa-can-break-trillions-of-encrypted-web-and-vpn-connections/ https://arstechnica.com/security/2015/10/how-the-nsa-can-bre... [3] https://blog.cryptographyengineering.com/2013/12/03/how-does-nsa-break-ssl/ https://blog.cryptographyengineering.com/2013/12/03/how-does... And so on and so forth. Also keep in mind that they are retaining a full take online-- storing all traffic on the internet. Even if it can't be decrypted in realtime, they can come back to it in, say 10 years, with all the advancements in brute forcing and potentially quantum computing available in 2027, to take a look at your 2017 activity. The emails you wrote, the websites you visited, all your chats, all your searches, saved forever.
- willstrafach 10y ago
  There is no evidence indicating that anyone can break encryption for mail servers using STARTTLS.
  - Karunamon 10y ago
    Not right now, no. But if you start saving encrypted streams, and then come back to them 10 years later when, perhaps, a TLS vulnerability is found...
    - willstrafach 10y ago
      I don't know what the media coverage stated, but the source material indicates that full-take is only stored in a 5 day rolling buffer. I believe 30-day for certain flagged selectors. Not at all indefinite, that is not possible.
      - hendersoon 10y ago
        As of 2011, GCHQ stored their full-take for 3 days and metadata for 30 days. Do you believe they haven't increased capacity in the past 6 years? Again, the discussion comes down to evidence versus reasoned speculation. Evidence of intelligence agency activity is generally pretty tough to come by.

ocdtrekkie 10y ago

> They wave away violating the 4th amendment of the US constitution by saying that copying, storing, indexing, and searching through US citizens' private information is OK as long as a human doesn't read it. This is funny because this is the same technicality companies like Google use to claim they aren't violating your privacy: That the only thing reading your email is their data mining algorithms, not humans.

srean 10y ago
That's a strained analogy. Its far easier to not use Google. I have to actively agree to Google's terms and conditions. OTOH I have to actively opt out (if at all that is possible). Ideally in a democratic forum I would have some say in it, but that exists only as a fiction.
- sokoloff 10y ago
  How many emails do you get from / send to gmail users? For me, it would be difficult to avoid using google because so many of my correspondents use it.
  - srean 10y ago
    I am saying that legally disowning your country, extricating yourself out of NSA coverage etc would be more difficult.

greydius 10y ago

I suppose it's a good time to invest in hard drive manufacturers.

Spooky23 10y ago

You need to separate national security activity from traditional law enforcement activity.

givinguflac 10y ago
Why? Parallel construction proves that law enforcement does not.
- cptskippy 10y ago
  Does not what?
  - hendersoon 10y ago
    They don't separate national security from law enforcement. They _say_ they do, but they don't. Parallel construction is when an intelligence agency tips off law enforcement about some nefarious actor (like say, Ross Ulbricht) and then LE has to construct a plausible scenario where their hard-boiled detectives tracked down the bad guy through traditional police work, not through spying on American citizens. [1] https://www.washingtonpost.com/news/the-switch/wp/2013/08/05/the-nsa-is-giving-your-phone-records-to-the-dea-and-the-dea-is-covering-it-up/ https://www.washingtonpost.com/news/the-switch/wp/2013/08/05...

willstrafach 10y ago

> match a search query on a nefarious keyword. Sorry this is not true. You are referring to an email chain involving FI maybe, and within that USPI would be minimized so that the analyst cannot see who the US one is. They would need to specifically override and would need a good reason to get USPI as that is taken seriously. > "Full take" sounds incredible-- it entails recording and retaining EVERYTHING passing through the internet. This is also not true. It sounded nice during the leaks and everyone loved to hear it, but this is simply not true, not plausible, and would be a huge waste of money because they would wind up with too much USPI (totally useless). The collection systems are mostly going to be getting traffic from MENA area, although the leaked information was vague regarding specifics, for obvious reasons. (No argument with regards to ethics of bulk SIGINT collection, many opinions out there on that, just commenting as it is very unfortunate to see falsehoods repeated due to reporters poorly understanding the source material)

elbii 10y ago

> This bill is great in that email stored on your ISP's (or Google's) servers is treated identically to that on your computer at home-- they need to get a warrant. That's a great step forward. It isn't BS. Except for one major difference: Gag orders. When a warrant is served to Google to access your email, you may never know about it. If they show up at your home and take your email server, who you gonna gag? Unreasonable or overly-broad gag orders are one reason I always advocate running your own services whenever possible (granted, this needs to be easier).

schoen 10y ago
The distinction you mention is real, although not all warrants against providers come with gag orders, and some warrants against people's homes authorize surreptitious searches, such as by breaking into the home when the target is away. (This was done for a personal computer as early as 1999 in the Scarfo case.) Lately law enforcement has been seeking warrants to hack people's computers and some of them have been granted; typically those have also been surreptitious searches. I think this is the greatest risk for surreptitious government access to self-hosted information, because some agencies have been getting pretty excited about this and there are a lot of contractors who will supply them with vulnerabilities and tools even if they don't have the expertise to develop those themselves.
hendersoon 10y ago
That is certainly true. The difference is that before they didn't need a warrant at all if your cloud-stored data was over 180 days old. A court order alone was sufficient.

gnopgnip 10y ago

The legality of this is being litigated now. https://www.eff.org/cases/jewel https://www.eff.org/cases/jewel

loeg 10y ago

> It's very reasonable to assume that everything, EVERYTHING, you do on the internet is recorded and stored by the government. This bill does not address that. I mean, that's hard to believe. I work for a large storage vendor. If this was true, the feds would be our largest customer by far. They aren't. I don't think SE Asia makes enough disk to store everything indefinitely, and given exponential network traffic growth, even storing the most recent few years is a problem.

dv_dt 10y ago
I might consider that they're just not buying it from your company, or you aren't privy to the transactions to those gov't customers. How much storage does 1.5 billion buy? (Ok it's not all just storage...). https://en.wikipedia.org/wiki/Utah_Data_Center https://en.wikipedia.org/wiki/Utah_Data_Center The explosion of data you mention is a real problem; though thinking about it, one might manage with a mix of light filtering, leaning on organizations that already store the relevant data, etc.
pbasista 10y ago
That is just one opinion against another. With little fact to support any side. In your comment, for example, you seem to suppose that those agencies do not buy storage on secondary markets. Can you confirm that? Because if they do, then your sales statistics are almost meaningless in this regard. Also, it seems like you assume that everything the agencies collect is stored "as it was transferred" (i.e. if the same YouTube video was viewed millions of times, they would store it in millions of copies). I think it is safe to assume that these agencies are a little bit smarter than that and if they store everything, they have some kind of compression in place to minimize redundant storage of the same chunks of data. In any way, I think that until someone like Snowden speaks up again and brings evidence from inside of these agencies, we are only left with speculation.
- hendersoon 10y ago
  There's no speculation on the UK. We know the GCHQ does this. Of course they generate quite a bit less traffic than the USA, given their smaller population. We don't know whether the NSA does "Full Take" today. That's speculation. I would be astonished if they didn't, and it is reasonable to behave as if they do. It's possible, it was possible 4 years ago, and storage is not getting _more_ expensive over time.

ThatGeoGuy 10y ago

> That's why it's absolutely critical that everything you do is end-to-end encrypted. These government agencies have the resources to crack most encryption, too [3] but at least you aren't making it trivial. This comment in particular seems to imply that government agencies can break e2e encryption. However, other sources would suggest that's not entirely the case. Snowden, for example, still seems to believe PGP works. I don't have reason to believe that it doesn't, or that the NSA can currently crack any arbitrary 4096-bit RSA keys. Further, there's a lot to be said about other modern encryption standards (ed25519 curves, forward secrecy and asymmetric double-ratchet protocols like Signal), which I don't believe have been shown to be compromised. I understand the importance of e2e encryption, but it seems that you're overstating the governments ability to crack it. I agree that SSL and TLS don't have the best history, and maybe we shouldn't rely on them, but it is also important to know that there are solutions out there which we still have quite a bit of reason to believe we can trust.

jwtadvice 10y ago
> This comment in particular seems to imply that government agencies can break e2e encryption. Much E2E encryption can be broken by government agencies. For HTTPS for instance, the government is a certificate authority accepted by your browser - as well as there having been multiple cases of collaboration with well known CAs to have certs minted for them. In this case there's no need to break RSA. In any case, much of TLS is open to various kinds of breakage (that is not as hard as RSA), and NSA docs showed a number of instances where intelligence agencies would sit behind load balancers and at data center chokepoints in domestic companies to gather information where PKI no longer poses a problem (behind the load balancers being referred to here as "end-to-end"). Basically, there's many ways to thwart "e2e" that isn't cracking 4096 bit RSA. Intelligence agencies do all of them.
- hendersoon 10y ago
  Yes, very much this. Full circle back to the famous xkcd comic that's always posted in HN articles like this. https://xkcd.com/538/ https://xkcd.com/538/
- 10y ago