The way to have transparency into politician’s communications is to require them by law to be made public, and to use law enforcement to make sure that this actually happens. It seems that relying on information going over email (as opposed to eg signal), and getting hacked (perhaps you want it all hacked, perhaps you are more happy while it is the side you don’t like getting hacked, either way I think one must acknowledge that by focusing on what is hacked, one is granting those hackers great control of the narrative) is not really very useful.
What I think is most shocking in this age of political hacks and leaks is the fact that people are outraged by it when it’s their side. Sure, the timing can be unfortunate when it harms their chances of re-election, but I’m surprised that I hear more about that, and calling it election interference, than I do about the actual contents of the leaks. Don’t like it when your side’s dirty laundry hurts their campaign? Solution: nominate candidates with less dirty laundry.
Email and associated protocols apply to everyone, public-individual or private. The same technology works whether you're a politician or an ex-girlfriend.
I also agree with your parent, if there is something we need politicians to do, it needs to be a law that makes explicit what the intended outcome is, rather than hold up an unintended consequence of a protocol feature as "good enough", especially when there's potential for collateral damage.
In fact, none of the examples in the article were from people up for election.
This has nothing to do with “blackmailing public officials” and for you to imply that I have such a desire is both uncivil of you to say here and says a lot about your world view. Blackmail is when you use evidence of illegal activity in order to coerce someone to do something against their will. Transparency and audit ability of our public servants is not blackmail.
One key flaw in your argument is that you seek a system that works, "even if they have gone rogue and used non-government-approved communication channels", which, it should be plain to see, absolutely does not apply to DKIM.
I perhaps read a bit too much into your statement, "Blackmail related to embarrassing sexual proclivities or anything like that is unfortunate, but kindly asking politicians to be transparent isn’t a realistic answer."
"Everybody who uses Gmail" includes a lot more private citizens than public officials.
For local politics this would expose the haggling/threats/backdowns not good for image making very hard to make deals
For international, how do you expect this to work when a politician is getting briefing Or guidances or heads up about dealing with an dictatorship or a unfriendly global power or an foreign company ?
Perhaps have a classification system ? Then everything will be secret classification
I don't think Matthew Green is arguing against transparency. What he's observing is that non-repudiation is an unintentional byproduct of DKIM's design. Because it's a byproduct, DKIM's users have made implementation decisions that make it susceptible to weaknesses in the unintentional non-repudiation property.
By 2030, a motivated nation state will probably have the ability to crack the 2048-bit RSA keys that Google is currently using for DKIM. Do you really want someone in 2031 to be able to contrive fake signatures for the emails of politicians in 2021?
I don't want a nation-state to be able to contrive fake historical signatures for my own emails.
DKIM without rotation and disclosure provides the capacity to do so with cryptographically provable integrity. Green's paper lists instances in which this has happened (as a proof-of-concept demmostration of the risk), and may have happened.
DKIM key rotation and public key disclosure at least denies adversaries this.
How could it be used for that purpose then if it’s proven to be unreliable?
It would seem that there’s more to gain in the short-term by those that have hacked Gmail accounts by exposing this, so it seems disingenuous, which you have to know, so it seems like people are fake-goading Google, causing others to actually goad Google, maybe to try to expose those that have hacked Gmail...
Pretty sneaky sis!
The distinction is in removing any doubt as to reliability.
Except that we're talking about leaks of emails that date back by years: the earliest Podesta emails are from 2010, back when Google was using 512-bit (!) keys for DKIM. Those were leaked in 2016, at which point 1024-bit keys were already considered crackable by a motivated attacker.
This isn't to say that those emails were faked, only that "an email written in 2020 that's verifiable in 2020" is not the target of interest.
The DKIM signature is proof only that whoever signed the email possessed the key, nothing more, nothing less. This, in turn, is a suggestion about the identity of the signer and possibly the author - but not proof.
Did DKIM change anything about the podesta emails? Or were they basically acknowledged as authentic regardless, and had a lot of other verifyable info in them?
Both journalists and investigative groups (and conspiracy theorists) treat DKIM as a sign of authenticity, even when the key material is long past its prime. Wikileaks still prominently displays a "verified" marker next to their archives.
> Did DKIM change anything about the podesta emails? Or were they basically acknowledged as authentic regardless, and had a lot of other verifyable info in them?
That's hard to say, but it's also not the point. The point with being able to crack the key is that a motivated party could intersperse false information with otherwise verifiable information. And, well, what's a conspiracy theorist to do? Only believe the non-juicy parts?
Yes, journalists verified it. But they consider it supporting data, just as they wouldn’t automatically ignore any email that had no DKIM signature.
Phone calls are never authenticated. Does anyone automatically believe or disbelieve recorded phone calls?
I mean, “conspiracy theorists” (in the common usage of that terms) already believe only what they want to believe.
I think the point boils down to expectation management: journalists (and ...) barely understand non-repudiation, much less why each of the following scenarios pans out:
* 2006 email + 512-bit RSA, leaked in 2006: probably authentic
* 2008 email + 512-bit RSA, leaked in 2012: potentially inauthentic
* 2008 email + 1024-bit RSA, leaked in 2008: probably authentic
* 2008 email + 1024-bit RSA, leaked in 2016: potentially inauthentic
...and so on. In sum: we're making life harder for the people doing real investigative work (since they're not technical), and we're giving fodder to the people who want to conspiracize. All because we're using a spam mitigation technique to provide properties that it was never intended to provide.
The wrong solution is to make previously private keys public to make any reasoning about past data impossible in the name of “hut journalists might get a wrong impression”
And yet, the deepfake equivalent to the suggested solution is one of "start showing deepfake as news" or "stop showing any video as news", neither of which anyone would consider a reasonable response to deepfakes. I just don't understand how DKIM is suddenly so revered as truth when almost no one knows what it is.
The point is that DKIM can be abused to lend undue credibility to falsified data... not that it can credibly attest true data.
These is absolutely no way you're going be able to educate the general public on the nuances of this. I mean, there are lots and lots of people who doubt the efficacy of vaccines and masks...
So can deep fakes. What makes deep fakes explainable and DKIM unexplainable?
If the journalists interests do not align about DKIM, how come they align about deepfakes?
I'm not saying journalists have any integrity. I'm just wondering why specifically for DKIM a "throw the baby out with the bathwater" solution is advocated, whereas for things like deep fake it isn't -- where the underlying truth is the same: "You can't trust what you see/hear".
Regardless, the fact that deep-fakes exist has absolutely no impact on whether DKIM has problems or not.
I have pointed out that in a similar case (potentially fake evidence), same actors (journalists) seem to have completely different incentives than those you hold so self-evident and I ask for an explanation of the difference - why is it so self evident that journalists have an incentive to not understand DKIM and not inform about it, but the same is not true of another concurrent challenge to evidence authenticity.
To me it sounds like you’re saying “journalists eat cotton candy because they like sweets, but they don’t like chocolate because they care about their teeth”. They might have this preference among cotton candy and chocolate, but the explanation is inconsistent and likely wrong.
By this logic, what the article is arguing for is to bring that same truth today: if DKIM no longer offers the same guarantees, but people think that it does, than it can trivially be used to forge emails that people will then wrongly trust, which is obviously worse than the status quo.
Of course, the more likely result is what the article suggests - if the scheme can be defeated, people will stop trusting it, and there will be no chance of forgery (at least not for very much longer).
He mentions the politicians because those were high profile cases. This could be used against anybody, not just politicians.
> It seems to me that especially when an elected official has something they don't want others to know about that it should be public knowledge.
Is this true of everybody else as well? Should anybody be able to deny an email they sent in the past? If so, we have to take this step.
People change over time and normal human communications have a natural sunset built in as people forget exactly who said what.
It's true, but I'm not sure it's as good thing as you believe. I was born in communism, and then later I lived through the transition and have seen many people use this exact mechanism that you mention to whitewash their biographies. People just don't remember long, and thanks to that all of the sudden everyone was a victim of the regime who fought for democracy, while in fact they were exactly the opposite. Many bad people not just got away, but also gain significant benefits thanks to "people forget exactly who said what" and it did a lot of damage to my country and the society. So, while people do change over time, and we all sometimes have said something stupid that we didn't really mean, IMHO as adults we all should stand behind the things that we say and hold accountable to at least some level for it.
And to protect people from other's misusing their past, perhaps it would be more beneficial to educate the crowd not to be overly judgmental and not to jump to conclusions like everyone on soc. medias just loves to do - rather than forcing individuals to lie about their past to defend of blackmailers.
Stuff like revenge porn already exists. Let's not make the problem worse.
Dear Ivanhoe,
I regret to inform you that your HIV test came back positive. Please contact my office at your earliest convenience to arrange a follow up.
Sincerely, Your doctor
"Hey Ivanhoe, your buddy from government here. That thing that we discussed, no problem I arranged everything, T says it's cool, just wire us the money and the project is yours."
In my view, if someone is going to blackmail me for some sensitive topic like being HIV positive or dox me in revenge, solution is not that I have to go public and lie that it's not real (and risk to be counter-proven it is) - but to have police put their blackmailing asses in the jail. That's the type of protection of my freedom and privacy that I hope for.
And in the end, who has ever believed people doing public denials? Once the word gets out, by the time you publish the rebuttal majority of folks will already have an opinion on it and that will stick with you for long time no matter what you say later.
Email is global. You and I are in privileged positions regarding access to capable law enforcement. We're also privileged with what our societies deems acceptable. We are the exception, not the rule.
If you're only thinking about how it affects you and what remedies you would have, then you clearly aren't looking at the big picture.
>the solution is [...] to have police put their blackmailing asses in the jail.
But now, you're saying you don't have meaningful access to law enforcement (in this context). So, why did you suggest a solution you know isn't viable? I don't get it.
To my mind, you've just made a strong argument for publishing DKIM keys since you readily admit law enforcement cannot tackle the blackmail problem. Indeed, even in countries with "good" law enforcement, they can't reasonably tackle it since the blackmailers almost always come from overseas (or are un-traceable).
B) Even though it's not viable for me to do anything to someone in Russia or China or even US for leaking my data, I see that as the only proper way to address this type of situations. If it's not possible now, then we should concentrate on fixing it and making it possible, instead of trying to lessen the impact, but at the same time helping those same blackmailers to easier hide their own steps (and a bunch of other shady characters who'd rather not be linked to their emails, from pedophiles to corrupted politicians). And also I don't see denying as a reasonable move here, as it comes down to basically lying publicly about the origin of your data and can just get you deeper in the trouble, especially if you're in any sensitive position and there're people out there actively looking to dig your dirt. AFAIK all PR handbooks on damage control say the same.
Even if this example is imperfect, it's really not that hard to imagine a scenario where some type of compromising information is sent to you. Perhaps even accidentally.
In fact, it does not authenticate any emails without a corresponding public key currently published to DNS. It provides specifically for "empty" or revoked keys to avoid such retro-validation.
Seems the central thesis is that because these messages are patently no longer authenticated by DKIM, we should eliminate any remaining hope of them being construed as authenticated by DKIM.
For example, I have servers that DKIM sign emails. If a person uses my servers to send a death threat, the FBI is going to want web access logs and smtp logs.
Or factored - Debian had a bug 12 years ago that caused weak SSH keys, a similar thing could happen to DKIM key generation (or has happened, but not yet discovered).
Some study showed many RSA keys in the wild had a common factor. A weakness of this family might be discovered with DKIM keys.
It is supporting circumstantial evidence, not proof of identity.
... for what they actually did.
You think the solution is allowing people to be blackmailed or otherwise publicly embarrassed for things they didn't do, while removing their ability to verify that they didn't do them?
Being gay is not a crime, and yet people can be blackmailed with it. It is very easy to open yourself up to blackmail by perfectly legitimate activities.
We shouldn’t be building technical systems that “trap” people, just because they might be doing something bad and might want to prove that one day.
Additionally you’re also ignoring the whole “people have the right, to not have their emails stolen” argument. DKIM signatures are only useful if the emails are stolen, are you trying to suggest that it’s ok to steal emails from people if they’re bad?
No, just the opposite, that is an excellent argument and I think that the privacy should be the real focus when we discuss the freedom, and not the accountability. Because freedom is not to be able to get away for the lack of evidence, freedom is not to put innocent people in that kind of situation in the first place.
Police state doesn't come from the ability to track citizens, it comes from the lack of transparency and government's misuse of the information. Now, reality is that having more data collecting increases the chances of misuse, but I think we're attacking the problem from the wrong side. Rather than killing the option to track emails, there should be much more control and transparency on when and how that data can be collected and used.
Option 1: DKIM keys stay private... "That email was just a joke, I'm not really gay" Option 2: DKIM keys go public... "That email was just someone else's joke, I'm not really gay"
Not really a difference, and with option 2 you can't prove you didn't send it (as far as you can prove someone didn't crack 2048 bit RSA and use that power to concern themselves with your sex life).
Being able to prove a fascist dictator who was killing people for being gay, was secretly engaging in gay acts themselves, might help your cause of protecting gay people.
How?
If the keys were public, they could claim forgery. Regardless they could claim their account was hacked, but they couldn't deny the message was sent from their account.
You can't be blackmailed by someone who has no plausible evidence.
People get blackmailed, shamed, hurt and even killed over mere rumors, speculations and suspicions. As long as people believe in something (because something merely look plausible), there's no need for a fancy crypto to prove some machine sent some email. I'd dare to say most people don't even understand what cryptography is and what digital signatures really are (who signs what and what exactly this means).
I'm yet to hear a story of, let's say, a brave dissident who got out of jail because of cryptographic plausible deniability property making their oppressors unable to prove authenticity of some leaked or intercepted correspondence.
Some google employees have direct and indirect access to signing keys or writing emails. Not many, and they have good controls, but still many people with the ability to sign messages.
Not to mention a Trojan infiltration or account takeover, of which thousands (if not millions) a day occur.
The DKIM evidence is, for legal purposes, a good hint but far from proof.
I fail to see how admissibility or lack of it, in a court of law or of public opinion, has anything to do with DKIM+Hunter Biden. Can you elaborate?
I am not sure why the DKIM for all emails were not released, or why this did not catch more media coverage by other news organizations I consider more reliable (like NYT).
From your link:
> The only way the email could have been faked is if someone hacked into Google's servers, found the private key, and used it to reverse engineer the email's DKIM signature, Graham, said.
https://www.zdnet.com/article/google-fixes-major-gmail-bug-s... is from Aug 2020 and discusses an SPF/DMARC vulnerability that was in Google since forever (and though reported 4 months before public disclosure, was fixed only 7 hours after public disclosure). The last google DKIM bug I'm aware of was in 2012, so I can't counter the specific claim about DKIM with evidence, but the assertion that "the only way to spoof x is to hack and get the private key" is not any absolute truth.
(P.S: I have seen no denial nor confirmation about the authenticity of the Hunter Biden data - only claims of Russian involvement. Make of that what you will. The DKIM is circumstantial data until there is confirmation or denial - especially, as you say, it's not all released).
I am not insinuating any wrongdoing from anyone, just bringing it to your attention, as you claimed to not know about it.
But it does support my thesis that DKIM or no DKIM is not what gives (or doesn't give) any credence to the authenticity (or lack of it) -- here we have a high profile case, with DKIM validation (which a lot of people on this thread cleim "is considered proof by people who don't understand it") and it seems to make no difference even in the court of public opinion - those who accepted it, accpeted it without DKIM, and those who rejected it as russian disinformation, rejected it even with DKIM.
You really couldn't find any? Come on. Did you Google "DKIM Biden"?
I've read literally hundreds of pieces on the hunter biden laptop, about half of them from republican leaning outlets, (I try to keep a balanced diet....) and none of them mentioned DKIM validation.
(For the record: I don't live in the US, I don't watch television, but I do try to keep a balanced news diet)
Evidence was destroyed.
Which is very distinctly different from a passive act of not maintaining evidence of the origin of every single thing. Keep in mind that no data is altered - the equivalent of all collected samples remaining intact.
It's still just as possible to collect email logs, their contents do not magically dissappear. They would have to be actively manipulated by the party which holds the copy that would be provided to the police (either reported to them or confiscated, etc). That same party could already decide to delete the emails or strip signatures and then alter them.
Edit: Removed the word "literally" because it was incorrect and caused distraction from the actual argument.
Edit: comments like https://news.ycombinator.com/item?id=25066116 and https://news.ycombinator.com/item?id=25118713 and https://news.ycombinator.com/item?id=25067189 also break the site guidelines quite badly.
Would you mind taking a look at this explanation I posted a couple days ago? https://news.ycombinator.com/item?id=25130956 It is my attempt to explain why we don't want users to flame each other here, even when the other person is ignorant or wrong. The reason may be different than you think, in which case perhaps it will have some persuasive power for you. I hope so anyhow.
>It's not a valid reason to break HN's rules.
>That could certainly be.
>But there are many different ways to stick up for the truth
Whoa. Dang, I have to say, I feel a little slighted. I'm neither ignorant, nor wrong, and I'm aghast that you would insinuate that.
I've contributed faithfully to this site for a decade. The other commenter has -15 karma because, as you noticed, his comments are largely childish, combative and unsubstantive. It's embarrassing that you are validating him.
His claim was that Google publishing DKIM keys as described in the article would be "destruction of evidence", but that's provably untrue since there would be neither intent or willfull neglect on anyones part.
Literally (yes, I mean literally) no-one else here on HN, or anywhere on the internet, has legitimately attempted to argue this. It just doesn't hold up to basic scrutiny. "Destruction of evidence" is a very specific legal term with very specific meaning [0][1][2]. He seems to be distorting it in a Guilianni-esque fashion - "It's fraud! ....But no, your honor, not in the legal sense. More like in my own made-up imaginary sense!".
I've been restrained and as courteous as possible (under the circumstances), but even after you tried to squash the thread, and I stopped commenting, he's continued to insult me. You seem to be tolerating it.
I would have appreciated it if you enforced sanctions against obviously bad actors and remained completely neutral. That is what you're known for, but I respectful think you've failed in this case. At any rate, I know you have just about the hardest job on the internet, so I'll go ahead and chalk this up to misunderstanding.
[0] https://definitions.uslegal.com/d/destruction-of-evidence/ [1] https://www.criminaldefenselawyer.com/crime-penalties/federa... [2] https://en.wikipedia.org/wiki/Spoliation_of_evidence
This is a way of pre-empting the objection "But the other person is wrong and I'm right", which otherwise is the most common reaction to getting moderated. Since the moderation issue is about how people treat each other rather than how right or wrong they are, it's helpful to take it off the table in this way. If you think about it, it's a way of raising the bar for behavior on HN and in that respect is a stronger moderation reaction, not a weaker one.
I am careful, when using such phrases, never to actually take a side on the issue of rightness or wrongness. Remember that in every argument, the other person considers that they are the one who is right; moderating like this is a way of temporarily standing beside them from that perspective and pointing out that nevertheless, they should not have broken the rules. (An exception might be if I happen to personally know the truth about the point under dispute. But I know nothing about DKIM; I barely remember what it is.)
If you skim through https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que... you'll see that I use this approach all the time when trying to get people to follow the rules. It's true that I tried harder in this case than I usually do. That's for reasons which are out of scope of this thread, and had nothing to do with tolerating bad behavior (see https://news.ycombinator.com/item?id=25162543).
Dang, I thought you would have understood that.
If you'd be willing to take a look at https://news.ycombinator.com/item?id=25130956, I'd be curious to hear your reaction.
I can help. Fly out, I got a guest room. I guarantee you won't regret it.
All blackmail involves things a person actually did... otherwise it would be libel or slander.
You seem to be arguing that blackmail shouldn't be illegal.
No matter what you think about politicians, it is a failure of cryptography, or perhaps our common application of it, that the signatures we use to assure our conversation partner of our identity can also be used for our conversation partner (or divers third parties) to prove what we said.
Compare https://en.wikipedia.org/wiki/Off-the-Record_Messaging which solved this problem quite a few years ago. Off-the-Record Messaging allows your conversation partner to know that they are talking to the real you, but does not empower them to prove that to anyone else.
The purpose of DKIM is to help prevent spam, not to verify the authenticity of the sender.
Even google didn't bother to rotate their DKIM keys as recommended by the standard, so one wonders if the google keys are stored in a cage guarded by lasers and dogs or if there are copies on someones laptop somewhere and any sysadmin with a gambling problem or a secret affair could have leaked them to an unscrupulous journalist or a spy.
> This is an amazing resource for journalists (...) But it doesn’t benefit you.
If it's an amazing resource for journalists it benefits me.
I think his point is that the DKIM signatures could be used to verify that you did, in fact, send something worth being blackmailed over, rather than having the plausable deniability of saying that your DKIM private key from that period is already public and thus could be forged.
Which, to me, sounds similar to the classic XKCD "Theoretically, I use 2048bit RSA encryption and the hackers can't get my data. In Reality, they just beat me with a hammer until I give up the password." Maybe a public DKIM argument would hold up in court, but if we're just talking reputation blackmail among family and friends, it aint it chief.
It seems to me that especially when an elected official has something they don't want others to know about that it should be public knowledge.
After all an efficient marketplace only is efficient if all actors have access to as much information as possible.
EDIT: As a follow up, several people point out that it could happen to me or a family member, but this seems even further reason to have DKIM so that if someone attempts to blackmail me based on the contents of my email, checking the DKIM signature makes it even easier to disprove a bad blackmail attempt.