Rsync.net warrant canary
rsync.net
rsync.net
1: https://en.wikipedia.org/wiki/Warrant_canary#Usage
2: https://web.archive.org/web/20131103121048/http:/groups.yaho...
edit: Yikes, remind me not to compliment people here. It boggles the mind how this could be seen as controversial. Did people assume sarcasm? I was being sincere. It is a good comment.
double edit: back up to 0! Suppose I sounded a bit sarcastic.
Warrant canaries were a reaction to the NSL process, which is invariably (and, I guess, permanently?) gagged, and was seen as exceptional and in some sense extra-judicial. It would be newsworthy for a service to be NSL'd, and further evidence of dragnet surveillance programs sweeping up Americans.
Ordinary search warrants and disclosure demands occur, presumably, all the time; they're issued by courts in individual felony cases, such as for drug conspiracies, child pornography, and white collar criminal conspiracies. Serious crime happens all the time; it's not really all that newsworthy for a warrant to issue in, like, an insider trading case.
So, what does it tell us if this particular canary document was taken down? Perhaps the DOJ is working with the intelligence community to dragnet the service, or establish a durable norm of being able to transactionally extract records that will amount to the same thing as a dragnet. Or, maybe, just some random state court judge in Oklahoma decided it was likely that somebody's meth distribution business kept records in that service. One of those is interesting, the other not.
Why not just have more than one canary if you're going to do it this way?
It's been this way for a long time, and I'm just now having this thought, so it's equally likely that my take here is just faulty; if so, let me know.
In fact, if they didn't inform the public of regular warrants, only (indirectly, via the warrant canary) when they were gagged, it might incentivize the government to only serve regular warrants instead of NSLs, which surely isn't the intention.
If regular warrants happen often, I guess they could make a separate mechanism to inform the public about them, but I guess it doesn't, so far? In any case, if they aren't gagging, they wouldn't need a canary to inform the public about them.
At the point where you're logging warrants, you're not doing a canary, you're doing a transparency report. The transparency reports we read are, as I understand it, the product of negotiations between companies and the DOJ.
If they are gagging, multiple warrant canaries might work, but I think it becomes harder to argue that you aren't communicating anything if you stop publishing one but still publish the others. Taken to the extreme, you could have many canaries, for example one for every user ("there has been no warrant for your information"), and users might even appreciate that, but stopping with publishing a specific subset of them can communicate a lot of information.
It also removes some of the plausible deniability ("maybe they just forgot to update the warrant canary") that comes with having only one, I think.
See my response to your parent, upthread.
The canary already contains a list of all warrants we have ever received - it just so happens that list is zero items long.
Some of the best discussion and counsel I have had took place during the EFF Canary Summit which was held at NYU in 2014. These disussions took place under Chatham House Rules, however, so I can't attribute it to anyone. Again, best characterized as "wildly divergent".
Of note:
rsync.net is, to a layperson, best described as a virtual safe deposit box. There are no abilities to publish, share or work collaboratively on data. In addition, an rsync.net account is relatively expensive. The kinds of activities that might attract a NSL or gagged order are naturally repulsed by these structural factors.
It's also a good filtering mechanism for non-technical clients, but I'm drifting off-topic now ...
It's not entirely clear from the canary itself but the idea is that we will list all (non secret) warrants in the signed message.
Currently, the PGP signed message starts with the line:
No warrants have ever been served to rsync.net, or rsync.net principals or employees. No searches or seizures of any kind have ever been performed on rsync.net assets, including:
... and if we were ever to receive a lawful warrant or notice, etc., we would then change that wording to reflect the date, service, location, etc. (in addition, presumably, to complying fully with this lawful order).
So the warrant canary would continue to be updated, on schedule, but with a new listing of all warrants served in the signed message.
On the other hand, if a NSL / secret / extrajudicial warrant were to be served we would be unsure of the correct course of action and would need to carefully consult our legal team, advisors, board of directors, etc. ... and that would take weeks.
The warrant canary would become stale as we performed our due diligence and retained and received proper legal counsel.
A normal warrant, or a "normal" order for search or seizure, etc., is not out of bounds in a free society - we will happily comply with such lawful orders which have proper jurisdiction, etc.
We will also log them, publicly, as is our right. It just so happens that as of today we have never received such a warrant or order so the canary contains zero of them.
However, if we receive a NSL or other such "gagged" orders then we will need to consult with our principals and legal counsel, directors, etc. and that takes time.
The canary will surely expire as we perform that due diligence.
See also, below, my characterization of the warrant canary not as a legal hack or nose-thumbing, but as a poison pill provision that we ingested 17 years ago.
Therefore it's a very easy call to stop publishing the canary, keep running our business as usual, and then update the canary with the relevant (probably non-personal) details of the warrant we received once the dust has settled.
Alternatively, our counsel, along with the jurisdiction involved, might find a satisfactory "we were served with a lawful warrant in blah blah District in Colorado" and nothing else that does not violate the seal.
One gigantic advantage (among many) of running a very tightly held, lean business firm with no debt is that you can be patient.
1. Receive an ordinary warrant in a criminal case, which will have non-disclosure language.
2. Take the canary down.
3. When, say, 6 months later, the case goes to trial and the warrant is unsealed, you'll put the canary back up, having logged the warrant.
They can’t be compelled to update it, because that would be compelling them to lie in real time to their customers and the general public.
National security can be used to make you stay silent, and cooperate by not hinting at secret material, but it can’t be used to force you to create and stand by a made up story.
No. The canary either stays up and is freshened every week or it stays up and doesn't get freshened.
Removing it is not part of the procedure at any point.
But if I was the gov, I would go to your webhost or colo provider first, the gag order would prevent them from telling the subject like rsync.net and the search warrant would allow access to read and tamper with code and data. This is why GPG signing is great, private keys won't be on devices you don't physically control if you do it right. And also why you shouldn't trust native encryption of cloud providers like iCloud where ultimately you are not managing the private key directly.
Lots of towns were setting up speed traps to catch speeding motorists, with very low speed limits, so the AA paid uniformed boys on bicycles to find the speed traps and flag down motorists to warn them before they reached the speed trap. This practice was legally challenged as obstructing the police, so the AA inverted the scheme. They instead paid boys on bicycles to salute motorists by default, except when there was a speed trap. If you saw the uniformed AA boy standing by the side of the road and he wasn't saluting you, you knew there was something wrong ahead.
The idea was that the law might be able to ban saluting to warn motorists of the police, but the law couldn't ban not saluting to warn motorists. The law couldn't compel a salute. Evidently this worked, because the AA kept up the practice for a few more decades, before eventually discontinuing the practice in the 60s.
The warrant canary is always thought of as a gimmick or a "legal hack" or a "smartypants" construct ... but this is the wrong way to think about it.
The warrant canary is a poison pill provision.
rsync.net is a real company. We have a proper board of directors and outside advisors. We have legal counsel. We have shareholders.
We could not possibly respond to any kind of warrant - extrajudicial or otherwise - without careful consultation with all of these stakeholders. That takes time and that means the canary will expire.
So it's wrong to think of this as an act of defiance at the time of service - that ship has long since sailed (2006).
Instead, it has shifted the landscape of warrant service, irrevocably, in advance.
That's not expression; that's like being told you have to put a notice on the front door of your business, and trying to avoid it by saying "I wasn't disagreeing with the notice, I just haven't had time to put it up". Yeah, they're going to shut your restaurant down.
I agree with your take across the thread that this is mostly moot because of the nature of your business; I think the discussion of whether warrant canaries work is interesting, but the question of whether rsync.net is going to have to resist government searches is not, like, a major controversy. You're just not going to get searches at all, because you're not that kind of company.
I do not see rsync saying that they will necessarily take down the canary when it expires. Just that they will not update it while they consult their lawyers and stakeholders.
> being told you have to put a notice on the front door of your business, and trying to avoid it
In the US, the case West Virginia State Board of Education v. Barnette established that the government cannot compel speech. That is the basis of the presumption that the government cannot force a company to update a warrant canary.
Cease updates of a canary is doing nothing.
Product packaging is part of commerce, which is regulated. Firms voluntarily engage in commerce.
Cease updates of a canary is doing nothing.
There may very well be reasons warrant canaries will hold up in court, but "canaries aren't commercial speech" seems very unlikely to be one of them.
(1) Courts initially make a decision about whether or not something constitutes "speech" versus a part of a business transaction or something like that. It's possible that a nutritional label isn't considered speech as much as it is product or a commercial transactional component or something. That doesn't mean the same arguments would or would not apply to a warrant canary.
(2) The food labeling analogy is generally forced transparency. So for example, the alternatives are to either not report nutritional information or to lie about it. The former is a legitimate concern I think from a free speech perspective but refusing to report contents might be reasonably argued to drift into the latter alternative, which is basically fraud. The government coerced speech is to move speech away from fraud.
Asking someone to continue posting a warrant canary basically amounts to coercing fraudulent speech, which is kind of tricky. It would be akin to the government compelling you to not only put nutritional labels on your food, but also to put false nutritional labels on your food. The government coercing continued posting of a warrant canary when a warrant has been served moves speech toward fraud, which seems really different to me from the case with nutritional labels.
rsync has posted here on the thread that they'd have to consult lawyers and whatnot about whether or not to continue posting the canary, and that might take time. I think this is more nontrivial than some people seem to give them credit for, in the sense that lying about it might induce various lawsuits about fraud from customers and stakeholders that have a a financial material interest in them being transparent.
I'm playing devil's advocate here and not trying to adopt any particular position. I think it's entirely possible a court could throw out the idea of the warrant canary on trivial grounds. But it seems to me coerced fraud is different in subtle but important ways from coerced silence.
There is no situation where we take the warrant canary down.
We either keep updating it or we don't.
There is no third option where we remove it. Or put it back later, etc.
It's either up and fresh or it's up and stale.
Again, I liken this to a poison pill which suggests the die has already been cast - the decision was made not now, in response to a warrant but in 2006.
If we do something - like remove the canary - that model starts to fall apart.
Doing something is interesting and provocative and actionable, etc.
But if we do nothing it fits the poison pill model very well.
The canaries are set up so whoever has the website stops the updating. Nothing is taken down. Nothing is updated.
The rsync.net Warrant Canary is now 15 years old - https://news.ycombinator.com/item?id=26960204 - April 2021 (13 comments)
Rsync.net Warrant Canary (2006) - https://news.ycombinator.com/item?id=5899197 - June 2013 (50 comments)
Rsync.net Warrant Canary - https://news.ycombinator.com/item?id=5837351 - June 2013 (1 comment)
Show HN: The rsync.net Warrant Canary - https://news.ycombinator.com/item?id=4834362 - Nov 2012 (1 comment)
Rsync.net Warrant Canary - https://news.ycombinator.com/item?id=702247 - July 2009 (46 comments)
Surely this convoluted scheme is still telling people you've received a warrant??
Has there been caselaw on this?
See: Apple vs FBI in San Bernardino.
It is important to note that the more "low-effort" style of warrant canary, or simply posting a static page that says that you have not been served a warrant, is probably not safe. Taking the page down is possibly an action you can be legally prevented from doing. However, that's not what rsync.net is doing. They are specifically posting a new one every week, because no court or authority can compel them to post one if they don't want to under US law.
Edit: My point is that "compelled speech" isn't the issue, it's "compelled false speech" vs "compelled silence", either of which infringes on the right to free speech in some way.
(That's not to say this argument, or any canary argument, would avail; who the hell knows, should be our watchwords in this matter.)
The FDA can't compel you to say anything you don't want to. But they can refuse to let you sell your product. The right to sell food isn't as strongly protected as the right to political expression.
> On Friday, the 9th U.S. Circuit Court of Appeals joined a slowly emerging consensus among the federal circuits, holding that governments have the right to mandate corporate speech “if the information in the disclosure is reasonably related to a substantial governmental interest and is purely factual.”
The latter requirement does not seem to be held in the case of compelling rsync to post an updated warrant canary claiming not to have received a warrant if they have.
Furthermore, in the case of cancer warnings, the actual law is phrased as:
> No person in the course of doing business shall knowingly and intentionally expose any individual to a chemical known to the state to cause cancer or reproductive toxicity without first giving clear and reasonable warning to such individual
so, it's not so much compelled speech "out of the blue", rather it's a requirement of doing business. There are other ways the business can comply with the law, such as not including the chemical in the product. In the case of warrant canaries, I'd be surprised if there's a similar law requiring them to be posted to do business.
I think the danger here is the interpretation of the prohibition on telling people you've been served with a secret warrant.
Whilst you seemingly can't force someone to do something the fact that not doing that thing is effectively telling people you've been served a warrant is grounds at least to take you to court.
I can only imagine this canary process being manual, if it were automated I'm unsure it could be considered speech (obvs: IANAL).
Until tested in court it's up in the area.
[0] - or at least I'm non-US
What isn't firmly tested in court is this specific application of the idea that the first amendment protects against compelled speech, although there have been other similar cases (see the Wiki article [1])
Source?
Edit, found this. Wonder why this reasoning is not used for false implications in all the other labeling for supplements, vitamins, herbal, and organic food.
Even for milk, there is “A2” milk sold next to regular milk implying that A2 is somehow more nutritious, but it is not required to say “A1 milk has not been shown to be any worse than A2”.
https://www.govinfo.gov/content/pkg/FR-1994-02-10/html/94-32...
> Because of the presence of natural bST in milk, no milk is ``bST-free,'' and a ``bST-free'' labeling statement would be false. Also, FDA is concerned that the term ``rbST free'' may imply a compositional difference between milk from treated and untreated cows rather than a difference in the way the milk is produced. Instead, the concept would better be formulated as ``from cows not treated with rbST'' or in other similar ways. However, even such a statement, which asserts that rbST has not been used in the production of the subject milk, has the potential to be misunderstood by consumers. Without proper context, such statements could be misleading. Such unqualified statements may imply that milk from untreated cows is safer or of higher quality than milk from treated cows. Such an implication would be false and misleading.
>FDA believes such misleading implications could best be avoided by the use of accompanying information that puts the statement in a proper context. Proper context could be achieved in a number of different ways. For example, accompanying the statement ``from cows not treated with rbST'' with the statement that ``No significant difference has been shown between milk derived from rbST-treated and non-rbST-treated cows'' would put the claim in proper context. Proper context could also be achieved by conveying the firm's reasons (other than safety or quality) for choosing not to use milk from cows treated with rbST, as long as the label is truthful and nonmisleading.
Also there is a difference between BST (bovine somatotropinnatural) and rBST (r stands for recombinant, the manufacturing process) the synthetic kind. The FDA's "concerns" are transparently based purely on the commercial interests of rBST producers like Monsanto and their customers'.
In my memory the comments used to be more positive about both the effectiveness of canaries and about the sheer cleverness, ingenuity and spirit of fighting back against tyranny.
I wondered what happened to the HN hive mind over the years, has it become more deferential to the authoritarian state, more accepting of the erosion of essential freedoms in exchange for ostensible safety, more dismissive of clever hacks?
But then I read a thread from 13 years ago: https://news.ycombinator.com/item?id=702247
While there's a noticeable positive sentiment, many of the skeptical comments of today could have been copied verbatim from July, 2009.
So perhaps my memory is faulty.
>But then I read a thread from 13 years ago: https://news.ycombinator.com/item?id=702247
>While there's a noticeable positive sentiment, many of the skeptical comments of today could have been copied verbatim from July, 2009.
The difference is that in 2009 and now a Democrat is in the White House, so those who are convinced that "their side" can't do anything wrong are more willing to accept the authoritarian state.
Look at the Warrant Canary Wikipedia page. Would Jessamyn West—associated so closely with Metafilter, a site that basically self-destructed because its hivemind became so self-reinforcing as to immediately shun anyone at any variance with it in a way that would make a Maoist struggle session conductor proud[1]—have created her sign for libraries if George W. Bush had not been president at the time? I doubt it.
Saying that you shouldn't rely on legally untested warrant canaries as a sure-fire protection is not the same as agreeing with the executive branch's overreach abetted by a supine & craven Congress and a Judicial branch all too inclined to defer to the executive when spurious security justifications are trotted out.
"Not updating a warrant canary" isn't the crime here. "Creating a warrant canary" is the crime, as it's an action whose only purpose is to violate a gag order.
Or that if they ever receive a gag order then their earlier action was retroactively illegal?
There is no gag order. That's the point.
Are you suggesting that any action that might violate any future order is a crime? If so, what do you base that on?
If not, since there is no order before them, are you suggesting there is a specific statute that they are violating? If so, which one?
If not, what would make this behavior criminal?
It walks a really fine line IMO. The way I see it is it works in two ways:
* When you create one, you're not referencing a current gag order, but rather an imaginary future one. If creating one is a crime then doing just about anything (deleting an email or throwing away a letter) is also a crime because you can be destroying evidence in a potential future case or investigation not brought yet.
* Once this warrant is issued, then your run into this case where any associated gag order is "one way". That is the government can order you not to disclose the warrant but they cannot compel you to lie and update the canary to reflect that you still have not gotten one yet.
INAL.
No that would be silly and there would be no way to prove malicious intent. You could make a million arguments for a canary in the same way.
setting up warrant canaries is probably a good idea if you have the ideological conviction to face the government's bullying. spending some money on getting lawyers' opinion is also probably a very good idea.
but all this serves as evidence for intent.
so it's not that simple to do it with some persuasive/plausible alternative explanation.
the whole point of gag orders and the mandated/compelled performance is to help whatever investigation. of course if someone doesn't cooperate with law enforcement and/or the prosecutor that's their choice, and in this case it's basically itself a charge. (obstruction of justice is not a new thing after all.) the complication is that speech is a special action.
There are subtle and/or complicated reasons to believe the canary would work! I'm not here to make the case that canaries are invalid, only to establish that among legal experts, this is not a settled issue.
Has the supreme court actually ruled on the constitutionality of gag orders? Until then it's not completely settled that it's despite rulings in lower courts.
Specific gag orders can and have failed in appeals!
The problem with this argument as it pertains to warrant canaries is that defeating the gag order also defeats the purpose of the warrant canary. The question we're begging here is: if the gag order survives strict scrutiny, why won't the order to keep the canary up?
The State SCs I have researched are the same; they choose if they want to hear the case. If they decide to not hear it, the cert is denied which doesn't mean the argument was decided. It just wasn't heard by the court.
Judiciary Act of 1925
https://www.uscourts.gov/about-federal-courts/educational-re...
The issue is that people post a warrant canary before they have any warrant served, so it'd be a strange situation of being legal before receiving a warrant and then suddenly transitioning to illegal. Maybe one way round it would be to make it illegal to change any procedures upon receiving a warrant so that they would be forced to continue posting warrant canaries that are then false.
It's not simple and it isn't resolved in law.
I was addressing the parent's comment about how warrant canaries should be illegal due to them being an attempt to get around the gag order, but doesn't that involve a gag order making a previously legal action into an illegal action and assumes knowledge that a gag order will be issued in the future?
The bottom line to me is, encrypt your data before it leaves your control, and cross your fingers that whatever tool you used did it correctly. If the government wants to see which sites you are connecting to, it's easy enough for them to just ask Spectrum, Comcast, etc. So IMO, warrant canaries are useless.
Where "cross your fingers" is equivalent to the "warrant canary has not been revoked" when encryption is involved.
That seems useful. Thanks!
Why is every user of a service supposed to care about something for it to have value in the service? Why is everyone supposed to agree on what level of connectivity obfuscation makes them feel comfortable? Why can't someone find value in simply knowing the government isn't monitoring how they use services without notice? If it has no impact on monitoring ability why would the government bother getting a warrant in the first place? Why can't someone be interested in knowing how pervasive invisible warrant requests are? Why should everyone equally be comfortable with just encrypting and calling it a day?
Even though I don't really have any direct utility from a warrant canary (e.g. Reddit's warrant canary went away ~7 years ago and it didn't trigger any direct consequences with me using Reddit) I still find them useful and, knowing how relatively lax I am about privacy/security compared to many, I have no doubts many find direct use of them. For some other specific companies/services/projects though I could even see a warrant canary possibly having high direct value.
I guess that's what I was asking, is why or how? I didn't say they don't have value as a fact, I said I don't understand the value. Why are they useful to you, ie, what would you do with this knowledge if your storage service had a warrant canary and stopped updating it?
It really depends on how you use rsync.net.
The stated purpose of our product is: An empty UNIX filesystem to do anything you'd like with.
So ... if you are using a sophisticated tool like 'borg'[1] or 'restic' or 'duplicity' then no action would be necessary. Your data at rsync.net is garbage ciphertext that we do not hold a key to.
On the other hand, you may, as many thousands of people do, use rsync.net as a dead simple SFTP/SCP endpoint[2] and just copy files there. In this case your threat model and exposure is very different - but presumably you realize that.
Yes, you are thinking correctly when you describe your bottom line as:
"... encrypt your data before it leaves your control ..."
... and we encourage all of our customers to approach it this way.
Wild tangent follows. I promise it'll connect to the original thread.
From a certain perspective, all discussions about personal rights are silly. Almost every free-speech case started with someone being an asshole. Fourth Amendment search & seizure cases almost always defend a person caught red-handed. Bitcoin is evil because it's used only for drugs and prostitution. And so on. Most of us live in a world where we don't ever need to assert our rights. As long as we're not an asshole, we don't have strange religious convictions, we aren't a journalist trying to expose outlandishness, we don't commit crimes, etc., then we can live our whole lives as if the Bill of Rights didn't exist.
The problem with that observation is that the borders describing acceptable conduct are defined by something, and that something is our personal rights. If the Bill of Rights didn't exist, then "acceptable conduct" would surely be a smaller set of actions. We'd have less freedom, and we'd all miss the things we couldn't do.
So the fact that nearly all controversy about personal rights seems silly (including the "I don't care about privacy because I have nothing to hide!" argument) is extremely significant. If the average person thought these discussions were important, then society would already be feeling constrained by lack of freedom, and we'd be in bad shape -- dictatorship, surveillance state, etc. And if nobody discussed them at all, then we'd probably be in a state of chaos -- survival more important than freedom, etc. Having silly discussions about personal rights is probably right at the sweet spot. The only thing worse than believing Bitcoin is useless is living in a society where everyone believes it's essential!
This is why it's OK to feel like a warrant canary is silly. You should. But you should also take very seriously the fact that you're able to feel it's silly. The moment you feel warrant canaries are important, it's too late for our society.
(The word "silly" is mine, not yours, but I believe the sentiment is in the same neighborhood as your question.)
In fact, EFF even held a "Canary Summit" at NYU in 2014 which I was invited to. It was only held that one year, however, and once Apple and other large firms discontinued their canaries all of the steam was seemingly lost ...
[1] https://www.eff.org/deeplinks/2016/05/canary-watch-one-year-...
There is no known case of US charging someone over a warrant canary.
There have been plenty of warrant canaries that have ceased to be updated - e.g. Reddit, and more recently, Storj.
1. Forcing registered sex offenders to alert people in their neighborhood
2. Correcting misstatements
Often that is also often a condition of probation, meaning they are still under the authority of the courts and are still being "punished" by the system in leiu of prison. This is seen as the individual "giving up" some of their rights either by committing the offense or agreeing to the terms of probation instead of prison. This would also apply to #2.
Neither one of those justification for infringement of the 1st amendment would apply in the cases of an NSL which are already on very very shaky legal ground and gag orders on them have been ruled constitutional in the past, currently they are only constitutional because a person getting an NSL now as the ability to appeal the NSL to a federal court, something that was previously missing
First, freedom from compelled speech is not an inherently stronger (or weaker) freedom than freedom of speech. If the government can prevent you from saying something, then it can almost certainly prevent you from saying it by not not saying it.
Second, national security is one of the most powerful legal trump cards in practice. The government saying that something is necessary for national security will be treated as fact by the court, no matter how much evidence there is to the contrary.
Third, the purpose of freedom of speech is to protect freedom of expression. Speech that isn't expressive in nature has a much lower bar to clear for the government to be able to restrict it. Warrant canaries strike me as essentially commercial speech, which the government has pretty wide latitude to regulate.
Citation please
>> Speech that isn't expressive in nature has a much lower bar to clear for the government to be able to restrict it. Warrant canaries strike me as essentially commercial speech, which the government has pretty wide latitude to regulate.
This case law around NSL have not been vary favorable for the government, Appeals courts have struck down the gag order provisions of the laws in the place, and are poised to do so again should a case come before them. The current make up the Supreme Court also leads me to believe they would not look favorably on Gag orders, though they would on the larger issue of National Security
But what if they just put you under a gag order, and then when you stop updating the warrant canary they charge you with violating the gag order? Would that still fall under the compelled speech cases?
The former has been tested and is (for some reason) within the bounds of the first amendment. While forcing someone to publicly say something false almost certainly is outside the bounds of the first amendment.
Did we, in fact, create a poison pill provision in 2006 with regard to legal service, etc., ? Yes, we did.
Will this be a wild west data caper with dramatic conclusions and brave, desperate actions ? No, it won't be.
If there's one thing you should know about rsync.net it's that it's a very boring company. We're going to keep it that way.
If worst comes to worst, say "Sorry bro, lost the key in a boating accident, nothing I can do".
I wish every account had its own warrant canary.
I think an elegant way for rsync to handle this would be to just move the page from ".../canary.txt" to ".../canary-2.txt".
Sends the message that the first canary is dead, but leaves a second one up for the next warrant.
If we can't trust rsync.net to not do that then how can we trust them to actually stop producing canaries after the gag order? Is it something to do with how they are produced?
It is to prevent 'pregenerated' message. Also increases the entropy of the message vs just a simple date change.
> If we can't trust rsync.net to not do that then how can we trust them to actually stop producing canaries after the gag order?
Elsewhere in thread it was discussed that, legally compelling someone NOT TO say something is somewhat easier than compelling them TO say something.
As an interesting theory, if they were somehow 'compelled' by a court order to say they were OK for X weeks, they could just pregen the next X weeks without headlines, which would possibly be another way for the canary to squawk.
Closing stock price of multiple stocks.
This week's powerball numbers.
Count of babies born in several hospitals on a specific day.
The current nytimes headlines.
The most recent close of the largest stock markets.
The last power balls numbers.
Most of those would be valuable in their own right if you had foreknowledge.
I understand that in this scenario we trust rsync.net and are afraid of secret warrants. If rsync.net wanted to deceive is, they could do that by... just lying? What incentives do they have to pregenerate these messages?
I have high confidence that the individual tasked with this can quickly think of another unpredictable metric with which to datestamp the canary.
If the Feds really want to nail you -- to teach you a lesson and/or make an example out of you -- you're nailed.
"Democracy Versus The National Security State" by Marcus Raskin, 1976
The Supreme Court has repeatedly ruled on compelled speech [1], most colourfully on the pledge of allegiance [2]. It’s the theory Apple ran with in the San Bernardino case to avoid creating a decryption key [3]. A theory that was so successful the FBI pulled their case rather than risk a precedent for encryption.
[1] https://en.m.wikipedia.org/wiki/Compelled_speech
[2] https://en.m.wikipedia.org/wiki/West_Virginia_State_Board_of...
[3] https://en.m.wikipedia.org/wiki/FBI–Apple_encryption_dispute
There are supreme court precedents finding that the first amendment prohibits the government from compelling speech, which is the legal theory for why warrant canaries are not equivalent to directly disclosing the NSL's existence: https://en.wikipedia.org/wiki/Warrant_canary
If the government can compel you to hand over some information, well that's government. They can also injunct you not to tell people.
But, can they compel you to lie, publically ? Can they compel you to not publish such canaries - just in case they want to compel you later on?
And frankly, does it matter - is such a canary really just a form of social performance? If I don't want the US government to know something about me, I should stay the hell away from modern life, internet and so forth.
But the issue hasnt been settled in the courts. Twitter v Barr (2014) attempted to settle the issue, but the case was dismissed in 2020.
Edit: https://www.courtlistener.com/docket/4181259/twitter-inc-v-b...
*I say "in theory" because I understand that this was never tested in court and some big name law professors did not agree that this would stand up in a court of law. A judge would look at the spirit of the law that bans the warrant disclosure and consider that the entire setup of the warrant canary is aimed at breaking it. I tend to agree with this interpretation because any reasonable person would consider the whole design is aimed at breaking that law.
The point of gag orders is to keep a criminal from destroying evidence while an investigation is underway. Having an individual warrant canary for every customer would defeat this purpose. A general canary doesn’t appear to.
It would be a sad day for US justice if a judge ruled that every action was within the law but, because a reasonable person would conclude that the actions in aggregate were intended to circumvent a law, the defendant is guilty.
I would not assume anything about warrant canaries, and would expect the FBI et al to get the judge who gave them a gag order to also issue a warrant canary falsification order at the same time. Most organizations would not have the werewithal to resist, or the fortitude of a Judy Miller or James Risen and their willingness to go to jail rather than reveal their sources (yes, I know Judy Miller is a warmongering sorry excuse for a journalist, but her stand in this respect is admirable).
Also gag orders aren’t perpetual. If there were orders compelling businesses to lie, past examples would have come to light the same way we know gag orders exist.
This is the whole contention with Microsoft Office 365 being hosted in the EU and the CLOUD act.
More succinctly, the legal precedent would compel people to "break the law".
I'd argue such a conflicted legal outcome could only be resultant from unsound base principles.
Nobody can say, “you have this on your web site saying you have never received a warrant; here’s a warrant, and by the way, you must also continue saying (now lying) that you have never received a warrant.”
The government can tell you to /stop/ posting a daily or weekly notice. It can say “here’s a warrant; you may not tell anyone you have received this.” In that case, you would be lying to say you had /not/ received it, and discussing whether you have received it or not received it at all can be prohibited. In which case, you would be obligated to remove or cease updating any statements related to a warrant.
Although a warrant canary sometimes suggests a very principled party (e.g., I first heard of public librarians doing it, decades ago), at the same time, it seems probably counterproductive.
For one example, as a customer of a service provider, I want them to be stable -- not potentially antagonizing those who could shut them down, nor getting involved in what (to me) seem like ambiguous technicalities over what they can and can't do.
For another example, imagine you run a service in which you've committed to a warrant canary. But one day a warrant comes, and you realize it's gravely important for the canary not to die and tip off some genocidal warlord you didn't realize was a customer. Now you're violating the canary assurance to your other customers, which is an assurance that you should've anticipated you couldn't give.
Alternative: Some of the modern compliance reporting by tech companies, about warrants/censorship/etc. seems less likely to cause showstopper problems, can convey more info, and is ongoing rather than single-shot.
Of course there will be warrants and other compliances, for various jurisdictions, and conscientious periodic reporting seems to help with civic checks&balances.
(BTW, I really like the idea of rsync.net and its hard-working founder, have pointed new customers to them, and have a TODO to move some stuff to them myself. The only proviso I've mentioned to people thus far is that there's an unclear bus factor.)
The whole point of warrant canaries is that you can literally be compelled to silence under US law when dealing with intelligence organizations. It is impossible to state
"The FBI has requested our data, all our users reporting on police brutality might wanna get off our platform"
It might be legal to state "The FBI has never asked for our data", it might not, this hasn't been tested in court. They're not being hostile towards them, the CIA has literally arranged for multiple individuals to be assassinated, they should freak you out lol
0. https://www.documentcloud.org/documents/21046081-google-geof...
I'd guess in-house counsel could handle the day-to-day of this, looping in leadership when appropriate.
That home government might also be a resource (not an adversary), such as if a difficult request comes from another government, and advice or diplomatic assistance is needed.
No $5 wrenches nor soiled drawers necessary. No making assurances that you can't keep.
My opinion (beware, IANAL): by itself, stating this once, or as many times as one wants, is definitely legal. What's not tested is whether it's legal to give the promise to make such statements in the future with predictable periodicity.
If you mean reporting "we responded to X government subpoenas during this time," the whole point of a warrant canary is that the government can force you not to reveal that.
If your main concern is the instantaneous removal of the canary once they've actually been served an NSL or whatever, then a) that's assuming that such a canary is, in fact, removed the moment they respond to such a government action, and b) seems to be positing fairly exotic situations, which are unlikely to be pertinent for the vast majority of cases where such a canary is being used.
And especially don't do it while predicating your entire strategy on a legal technicality game.
a) If this is something that's going to be a problem, it'll be because the government has made itself your enemy, and
b) the government's inability to compel speech is more than just a "legal technicality game".
In the US at least the government can order you to not reveal the existence of a warrant. They cannot (in theory) compel you to state that you have not received any such warrants. Thus the canary.
I agree that being able to report on these activities transparently, even if it is done with a delay, would be preferable. But my understanding is that to accomplish that would require legislative changes to the current system.
...
Be skeptical. This is almost certainly just a marketing ploy so potential customers feel like their data is safe and secure.
On the surface, there appears to be no actual canary-specific value to keeping the text short. If you have a specific reason you disagree, feel free to bring that argument forward.
Norway: https://imageio.forbes.com/specials-images/imageserve/60ce8f... (2021 - the Norway curve looks like it was headed to 80% but forecasters adjusted it to be in tandem with the other nations.)
Tens of thousands view body of former Pope Benedict:
https://www.theguardian.com/world/2020/aug/03/former-pope-be... (2020)
North Korea's Kim sacks No. 2 military official
This one is funny. A bit more 'unpredictable' than having "NK's Kim launches missiles towards Sea of Japan" I suppose.
So my q, specially given AI, is are these 'current news' bits really unpredictable? And if the message is signed anyway (and we hope the key is not compromised) what other purpose does this serve beyond key rotation issues (and they don't rotate these pub keys, right?)
p.s. If these bits are supposed to be as unpredictable as possible, then we should note that any matter related to trends in industrial, technical, political, and major religious organizations (the Vatican) arenas are the bread and butter of security services of state actors. The current bits should be things that can neither be creations of state actors (i.e. sock puppets in 4chan starting a trend) nor matters that they by definition are laser focused (such as industrial output of near peers).
It's not supposed to guarantee Putin that the US hasn't seen his porn collection. It's supposed to tell me (a regular guy) that the government hasn't seen my vacation photos in secret.
The article you linked to is titled "Former pope Benedict XVI reported to be seriously ill" and predates his death by more than two years.
That isn't the purpose of the embedded headlines. They are to show that the canary file was not created and signed two years ago and only just now posted to the website.
They serve the same purpose as what used to be used in movies and TV years ago for "kidnapping" story lines where a photo of the kidnapped person, holding up a current copy of some major newspaper, was sent to the person who was being asked to pay the ransom (or perform some other action) in order to secure release of the kidnapped individual. They show that the item in question is current and not prepared well in advance.