Court rules encrypted email provider Tutanota must monitor messages
cyberscoop.com
cyberscoop.com
This ruling is a follow-up of a previous court order from 2020. We have posted a more detailed comment here: https://np.reddit.com/r/tutanota/comments/k3sfs5/in_englisch...
It simply didn't attract a lot of interest because most people figured that PGP-encrypted messages sent from throwaway webmail accounts was good enough.
Also the bitmessage software was really horrifically low-quality python, and the project seemed to take the bitcoin attitude that "the first implementation is the spec, multiple implementations equals bad" which was a turn-off. I had a really really hard time putting my faith in that codebase.
I think the idea behind bitmessage will resurface in a cleaner form if a compelling use ever arises. Enough smart people heard about bitmessage that the idea is now "out there" permanently.
If it ever does, I think a neat twist would be changing the Proof of Work (used to limit spam) into Proof of Storage like Chia uses, except that the storage "plots" are the active set of circulating bitmessages (which are encrypted and therefore have no exploitable structure, just like the hash chains in Chia plots) plus a smaller indexing structure. Two birds, one stone. This is not straightforward, but I believe it is possible.
>The decision will only impact unencrypted incoming and outgoing emails, as Tutanota can’t decrypt data that has already been encrypted, Tutanota added. It also said this should serve as a warning that for customers interested in maintaining their privacy, encryption is paramount.
That actually is a big deal, last I checked (in the US anyway) you're not really required to log much (if anything) as long as you can keep illegal content/abuse off your service.
1. Do KYC (Know Your Customer) checks to verify somebody's identity
2. Cross references customers to blacklists provided by the federal government (anti-money laundering laws, political sanctions, known fraudsters)
3. Maintain records of who you lent money to, when, for how much, and what the declared use of funds was
Etc, etc. There are a lot of reasons in a lot of industries where you're required to keep records of things.
I was just talking about messaging.
Given the sorts of proposals that have been made to effectively ban end-to-end encryption, the part where it only affects unencrypted messages (and possibly the ciphertext of encrypted emails) is the critical point. The court ruling requires the email provider to intercept emails they already had the technical ability to intercept, for specific accounts that have been connected to a specific crime by evidence.
> The decision comes as nations around the world have sought to weaken encryption to benefit law enforcement investigations.
When news articles talk about "an attack on encryption" it should be read as "an attack on companies being able to effectively employ encryption". While the NSA/CIA can and have put in backdoors and breaking changes into encryption standards, when leaders talk about how encryption hinders law enforcement they are trying to groom us for legislation (or executive pressure) which restricts businesses protecting their users and their data.
The FBI has applied pressure on Apple to not provide e2e encryption for icloud backups (https://www.theverge.com/2020/1/21/21075033/apple-icloud-end...) and I have no doubt other big companies have passed on stronger protection in the name of complying with government pressure.
https://www.ft.com/content/00e6b9d5-a9a2-488d-821f-de6813ca4...
His proposal is for "threshold cryptography" as a trade off between privacy and allowing law enforcement to do their thing when necessary: to break a message would require agreement / input from multiple parties, including the courts. I think potentially a good compromise.
If those public keys are on the client, it does not require the message to be sent to a trusted server.
mtgx's comment points to some practical implementation and logistics issues, though. Never mind how one would handle key revokations...
---
In case of the main article, though, that's irrelevant, as an e-mail provider necessarily has access to each unencrypted e-mail as they deliver them to and from the user. I can see it being the least bad way to log messages (if they would need to do so). Rather than logging in cleartext, or encrypting with a single key, it could be a threshold encryption where, say, more than one entity inside law enforcement/high court/etc are required in order to decrypt intercepted traffic.
I think our learned instinct has become to say "yeah, would be nice but good luck making anyone implement that"... But why not?
Isn't this just key escrow by another name?
For example the ability to open specific mails, specific conversations or with specific other parties under particular conditions, without revealing metadata of the rest. Or conditions like "if the person X communicated several times in a month with person of interest Y during the time crime Z occurred, open it up, but don't reveal the existence of anything outside that".
You can think of that as escrow of each message separately and just enough to match which messages you're looking for, while keeping the as much as possible about the others private. But actually doing it takes sophisticated cryptography still at the leading edge of research.
"their thing" being pointing a telescopic lens at a suspect as they enter their password, or planting a hardware keylogger when they leave their computer unattended?
Or is "their thing" more like mass surveillance of everyone?
Does Cheng feel car manufacturers have a responsibility to plant microphones in their vehicles, with "threshold encryption" keeping conversations in that car secret? What about construction companies? Law enforcement has no way to retroactively get a record of all conversations that took place in a room, even with a valid warrant. Does that make rooms beyond the reach of law enforcement?
What other everyday tools we consider "ours" should be required to betray us if whatever threshold he proposes is reached? And only the good governments will be able to reach these thresholds, right? If you look back through history, the risk of one's own government turning against them is so low, we should have no qualms about surrounding ourselves with tools ready to turn on us, should the state ask them to?
For a democratic system to function efficiently trust is essential. Citizens need to trust that a democratic government will not over reach beyond their stated power, and governments in power need to respect that citizen voters are not manipulable fools. And both needs to believe that the other can be held accountable, if necessary.
Without going into much detail, I believe that a decent compromise that can be reached on this issue is for the citizens to demand extremely strong privacy regulations and stringent wiretapping laws if the government wants us to forego encryption.
I'd rather not have to trust my government at all. I'm not sure that I buy the premise that trust is necessary- quite the opposite in fact. I think accountability is necessary. Any system that requires trust or goodwill is doomed to failure.
And I blame the citizens of that country when such things happen. Democracy survives only if you actively participate in it. A healthy democracy is where people are not apathetic and extremely cynical.
How can an email provider be required to spend time and money to spy on all of its users for a government without a specific warrant?
Edit: No Clipper chips. Not this again. No way. https://en.wikipedia.org/wiki/Clipper_chip
ISPs were required to save logs of their users' traffic in the EU. Eventually this got squashed by the courts, but that took years.
Beyond the Utah datacenter's storage abilities as a privacy eliminator and blackmail facility, there were a worse series of mass privacy-invasions in the US enabled by corporate interests including FAIRVIEW and the "Room 641A"s.
https://theintercept.com/2018/06/25/att-internet-nsa-spy-hub...
In these instance, it's not public what volunteering, complicity, inducements, cajoling, and/or leverage were used to attain such arrangements.
However, if the DNS server is located in Germany, they can force that instead.
Replacing the latter is particularly hard as DNSSEC operates a separate single trust anchor and is not as easy to circumvent silently as say x.509.
It is a strong case for using key pinning on all levels, from routing protocols through DNS all the way to email signatures.
Did you see anyone ban Turkey or China roots?
There is nothing anyone can do to decrypt the content, even if given access to mail servers.
As a side note: there has been complaints recently from the European Court of Justice that criticised that the German prosecutors are not independant enough for European standards. A reform is being discussed in German political circles, but nothing specific is to be expected before this years general election in September.
> its a local public prosecutor who orders such operations.
Unless Germany uses "public" in wierd way[0][1] the public prosecutor is part of the government. (It's debatable whether they count as part the national goverment, versus just the local/city/whereever government, but the emphasis on "German government" rather than "German government" suggests that's not what you meant.)
0: eg British "public schools", which IIUC are actually the subset of private schools that aren't limited to a specific profession/religion/other demographic
1: In which case maybe I'm just one of today's lucky 10'000[2] who gets to learn interesting facts about the German court system.
What is interesting about this is that the word "government" is usually translated into German as "Regierung", but the German term is more or less exclusively used for the leading part of a body politics (Merriam-Webster's definition 1). When refering to the combined whole of executive, legistlative and judicary we would typically use the word "der Staat" (the state) or would talk about the "Regierungssystem" ("system of government"), which is a bit inconsequently because the "Regierung" is only a part of the "Regierungssystem".
So I was reading the sentence of the OP more in the sense of the first definition, because this was closer to the literary translation into German, but can now see that the meaning of the word in English is less precise and therefore more context dependant.
Footnote 1 it is then; neat. I'm not sure what a more precise translation for that would be; "administration" (eg "the Bush administration") is the best I can come up with, but that's vague and ambiguous in other ways.
In what situations private emails of criminals are absolutely necessary? The only thing I can imagine is an exchange between two criminals of classified information for virtual currency.
Also, a lot of emails are from non-Tutanota providers to Tutanota. These emails are encrypted upon arrival with users’ keys, but they could be handed over to law enforcement before encryption.
Eh.
So the takeaway here is "If you're going to do something that law enforcement is likely to get a court order to monitor, don't sign up for an end-to-end encrypted email service and then send plaintext emails over it."
I don't think, as a society, we can preserve complete privacy of communication and also have a justice system that is seen to be just. If the bad people can trivially conceal all their communication, and yet we require evidence of wrongdoing in order to convict someone of a crime, how do we balance that?
Things like establishing intent (for murder vs manslaughter) have hinged on what the accused said prior to the act.
For white-collar crime, proving collusion has hinged on communication evidence.
If all the documents that prove fraud are stored encrypted with no way of getting to them, how do we prove fraud?
If we can't procure the evidence, how do we convict bad people of the bad things they've done?
Do we just prosecute the stupid people and allow anyone with an ounce of technical knowledge to do whatever they like?
Most evidence is not communications. Most crimes have a victim that can testify.
Otherwise, how did we ever prosecute crimes in a world were people just talked to one another in private?
Yes, there is usually evidence that is not electronic. But removing the electronic evidence makes it harder to secure a conviction.
In many cases, the victim's testimony is not enough. If it comes down to the victim's word against the accused's, then we come down on the side of the accused in these cases, because any doubt at all is enough not to convict. To change this we would need to change our entire legal system.
> Otherwise, how did we ever prosecute crimes in a world were people just talked to one another in private?
Communication is not just conversation. Letters and documents are also communication, and leave a paper trail that could be followed. Now, not so much. If all documents are e2e encrypted, we have a problem proving any crime that doesn't leave physical evidence.
Certainly, which is why those up to no good would make a point of not creating such a paper trail. If they had to communicate over an insecure medium they would use codes. It seems unreasonable to insist that everyone give up their normal privacy just on the hope that a criminal might fail to obscure their meaning on what everyone will know to be an insecure medium.
> does not consider itself a telecommunications service
Is there any point in reading further?
If they are not a telecommunications provider there is no law obliging them to do anything.
So it's all about legal definitions...
I'm sorry, I don't buy that the government should be recognized as having an absolute right to interpose itself in the middle of any two party conversation. Not even by warrant.
Justice systems are not, nor should be all-powerful, and that's the direction I see pop up more and more often to the point I'm becoming seriously concerned that rhetoric is sliding in a direction where liberty and freedom is sacrificed on the altar 9f public safety.
There is no point being free in a world where you are constantly micromanaged in the interests of maintaining the stability of the state and public, especially when those driving that edifice almost inevitably end up not aligning with the public they putatively serve, and consider anything less than half of the population an acceptable sacrifice to the whims of a slim majority.