Testing ISP Censorship
ryants writes "As part of a research project, Christian Ahlert ran an interesting experiment. He posted John Stuart Mill's On Liberty, which is clearly in the public domain, on different ISPs. He then sent the ISPs phony copyright violation notices. The results are troubling, with ISPs "acting as judge, jury and private investigator at the same time.""
FP censored by my ISP
While the fact that a UK ISP folded immediately, without further investigation, is somewhat troubling (mostly for customers of that provider), the size of the sample does not justify any sort of conclusion. I'd be really interested to know which ISPs have NTD policies, and which do not. The US ISP responded in precisely the manner that I would want my hosting service to employ. Exactly how widespread is the problem?
Of course, if we want this kind of behavior on the part to ISPs to stop, then they have to have some kind of legal shield from copyright damages while they investigate claims of copyright infringement. Normally, I would say that a reasonable approach would serve everyone well, but the legal system seems anything but reasonable these days, particularly in regard to intellectual property. So it would seem (if there is not such a shield already available) that there ought to be some kind of law or ruling that explicitly specifies the duty and liability of ISPs in the event that a copyright violation is alleged. Meaning, of course, that I draw pretty much the same conclusion as Ahlert does. I would prefer that he had documented the problem a little more thoroughly, however.
I remember your eyes, on the twelfth of July...
The actual results are presented in detailed form in a PDF file:
How 'Liberty' Disappeared from the Internet
Or see the Google text version.
This flagrant abuse of the "public domain" system has to stop. If people start reading and sharing intellectual property simply because it is no longer covered by any form of valid legal restrictions, how are people like myself supposed to continue to earn a living? Who's thinking about that?
Remember, EOB today! (Tick, tick, tick)
I am waiting for the day when people start sending(read: abusing) these kind of take down notices in order to shut down a site they don't like, or even worse, competition. Granted, he tested only 2 Internet Providers, but this could become a new form of DoS if providers just started shutting down sites without doing some research. The provider should follow up as the US provider did, give the site owner a chance to defend his/her work.
...It actually mentions a lot about doing just that and how easy it would be. Oh wait.. this is /. ... sorry.
Hmmm.
Let's see... So, he posted it to two, yes that's right--only two, ISPs. One of them took it down without looking into it, and the other actually researched it. So what's his answer to this "problem"? Government regulation, of course! Hmm... Does anyone else see a problem with both his experiment and his conclusion?
You probably shouldn't click this.
Yeah but you are also paying for that bandwidth. OTOH, if I were getting it for free then they can tell me what I can and cannot put up because in that case I am using their bandwidth that they are paying for.
Hmmm.
As the aleged copyright infringement was textual, the US ISP investigated. But put a piece of music or some open source software on a website and try your experiment again, this time citing the DMCA. I'll wager 10 out of 10 fold before you have time to check your email again.
The real problem is that ISPs don't want to be bothered. They don't want to get sued and have to pay a bunch of money to lawyers because of the actions of a couple of customers. So, when they receive a copyright infringement claim, there is very little chance they will actually investigate it, and even less chance they will fight the demand that the material be taken down. Instead, they will almost always automatically take down the "offending" material, no questions asked.
Exactly how widespread is the problem?
I doubt its very widespread, nor is it a "problem". ISPs are a dime a dozen. You can find them to pump your spam, serve your porn, email, whatever. If you don't agree to the terms of use for any reason, simply go to another one.
The slashdot crowd is so wierd about copyright infringements. If I owned an ISP and someone reported that there was a problem with copyrighted material on my equipment, I would take the stuff down too. How much time should I spend seeing if the stuff is a problem. I guestimate 0. If the material is that important and really is OK to redistribute, then a counter from the original person suplying the material will prove to me that its OK to put back up, and I would do it. However, aside from this sample size of 2 "research" experiment, I would guestimate that close to 0.0% of all complaints to and ISP about copyright violations are not copyright violations. I would imagine that most every ISP has a zero tolerance policy (aside from the sleezy ones that will host anything for a price) regarding copyrights, and don't care to spend any time figuring it out.
The problem is not that ISPs "acting as judge, jury and private investigator at the same time." . It's rather that they're not acting as jury nor a private investigator. They're not investigating wether the copyright is actually infringed or not.
6 years ago I was hired to track down websites that used my clients copyrighted pictures. It was never a problem to get the pictures removed. Of course, I could prove that my client had the copyright, but it would be a lenghty procedure. So, the ISP's took my word for it, thinking that if my claims where false, the person who uploaded the pictures would complain. I think it's the same reasoning we see today.
There is a clearly defined procedure in the DMCA for ISPs ... on receipt of the complaint, they MUST notify the site owner and MUST take down the material within a certain time.
On their part, the site owner MUST file a reply to the takedown within a certain time. When the ISP gets the reply, which is a legal document swearing to the ownership of the material, they MUST restore the material or allow its reposting (the ISP's part is over).
The complaining party (the claimant to the copyright) MUST at this point either file a formal copyright infringement suit in the federal court closest to the web site owner's place of business or shut up. Repeat complaints are NOT allowed.
Google reveals that the only web sites with sitebuilder/tandc.htm in their URLs are Freeserve/Wanadoo, and the pdf of the article describing the takedown. Therefore it must have been Freeserve or Wanadoo, who are the same company now. I'm glad I have started to move away from Freeserve, although most of my web site is still on their servers, I haven't migrated most of it yet.
But why private governance should be less intrusive than government regulation is something I have never quite understood. State censorship, while clearly problematic, is at least open to questioning and accountability. Notice and takedown is censorship without debate.
There are three differences:
1) Choice. Although you can question and debate state censorship, you have no options beyond that questioning and debate. If your government doesn't want you reading Mill's "On Liberty", your only real options are to defect or revolt. But for a private ISP, newspaper, radio station, etc, you have a choice. You have more than one ISP. You have more than one newspaper.
2) Guns. State censorship is backed up by guns, police, armies and navies. If they don't want you to publish a political screed, you're going to be in a world of hurt of you do. Private ISPs can't do anything but cancel your account. The MOST they can do is tell the state that you publishing a political screed, in which case this is an instance of state censorship anyway.
3) Ownership. Who owns the speech that should be free? In the case of state censorship, the state restricts your property of free speech. It has to leave its domain and intrude upon yours. But the situation is the opposite with private ISPs. The website is hosted on the ISPs property. It's their harddrive. They aren't intruding upon your property, but looking after their own. They never have to leave their private domain in order to restrict how their property is being used.
Don't blame me, I didn't vote for either of them!
I have two issues with this article.
First, the sample size is pathetic. Let's sample two ISPs, one in the UK and one in the US, and then draw overly broad conclusions based on those two results. Sounds good to me! (Or perhaps it sounds like a quick, insufficient study trying to support a premise rather than gathering evidence for one.)
Second, the article pays lip service to the issue as seen from the ISPs' point of view. (Granted I'm speaking entirely from the US side of things as I am unfamiliar with UK law in this regard.) ISPs have to deal with the DMCA. Okay, the author mentions this in passing. The author also, again in passing, mentions that ISPs carry liability for the information they host. Put the two together and what do you get? Take down first, look at it later.
However, the point is not that internet self-regulation is a bad form of regulation in itself. Instead, it is that the power an ISP has over content on the internet necessitates a clarification of the legal and governance framework. There ought to be rules about process. In other areas of governance it matters who governs and under what terms - this is also the case in borderless cyberspace.
There are rules and there is a process. It just happens that the author doesn't want to confront them dead-on. "ISPs have too much power, watch what happens when I threaten one!" just isn't a valid tactic in today's litigious society. That's akin to saying "Look at this! If I file a phoney report with the Police about my neighbor being a homicidal murderer, sometimes they come arrest him when they catch him holding a knife!" Flakey at best.
It'd be better off being summarized and turned into a Op/Ed piece. The paper itself seems very sound in structure, setting up methods and procedures for experimenting, developing a hypothesis and all the remaining factors for a good research paper. However, to have spent so much time in preparation for a sample of 2 is like spending a whole year prepping for a 10 second race.
Just my critical view of it, after several Associates, a Bachelors, Masters and starting on a second Masters degree.
Don't blame the isp, as a web host I have to worry about DMCA crap, RIAA, MPAA, piracy, lawyers, lawsuits, and litigation. After a while I simply adopt the attitude of "you guys figure this out, send me a copy of the court ruling, if it's OK to have it we'll put the site back, if it's not OK to have it then the site stays down".
No one wants to get caught in the middle of this crap, it's ALL too easy to make these bogus claims and it's ALL too easy to make GOOD on even bogus claims when things such as the DMCA and Patriot Act are out there to give censorship such a big stick.
I don't like censorship, but I don't like flinging my wallet at some sheister lawyer either. In the end I have to weigh the lesser of two evils and while I hate censorship I hate being bankrupt thanks to lawyers or imprisoned thanks to REALLY BAD LAWS even more.
--- www.f-theocean.com
I own the copyright to all messages posted by your user "hackstraw". Please delete his account and everything in it immediately.
Don't blame me; I'm never given mod points.
mod parent -1 hairsplitting...
So if you had, say, your masters' thesis, on that UK ISP and I wanted to cause you some trouble, I could sent them a bogus C&D letter and they'd yank your thesis offline without contacting you, or attempting to verify the legitimacy of my C&D letter, and you'd think that's a good thing for them to do?
Please THINK.
Lycos - on page 21, where he has shoddily neglected to blank it out. Ironic, as Lycos is actually owned by Terra, a Spanish (EU) ISP.
I agree that this study is a very good idea; I just wish he could have done it a bit better and more thoroughly. Two ISPs, one from the US, one from the EU, is simply not enough to draw any sort of conclusion.
In a way, I have to disagree with what the ISP's did in removing/blocking the content. I realize that the ISP's own the network as you've pointed out, but what would stop someone from makng claims about numerous sites over copyrighted material and having those sites shut down. The author doesn't do a very good job explaining the exact setup of the site or the details of the actions by the ISP, so I can't say with certainty that they acted improperly. If the ISP's just pulled the connection based on a bogus claim, imagine the havoc someone could play on web sites.
Picture the ISP for a local politicians website getting a notice that www.somepolitician.org was displaying copyrighted work. Picture that claim coming from www.running-against.org. Without some level of validation, claims could affect legitimate sites.
One would hope that ISP's would require some level of proof indicating the copyright infringement followed by a contacting the website operator to inquire if they can show rights to the use the content. I don't know if this would put too much burden on the ISP though. Maybe not the ideal solution, but seems better than just pulling the connection.
It's not just music. Ulink.net shut down part of my site when I was still hosted by them after receiving a complaint from the Beverly Hills police department for "displaying children with pornography". It was a background picture of the Trix rabbit saying "tits are for kids". Didn't bother telling me about either. It was shut down for days before I noticed and sent them an email asking what was going on.
7th Heaven sarcasm
Ulink quote:
We have been contacted by authorities regarding your displaying of a child star on a page with pornography. For your own protection we had to take down that page or you could be in serious trouble. In the future, please do not display pornography and children together. It is against the law and Ulink policies.
riding round the world on an old motorcycle
And look at counternotices on: this page
There is a formal procedure that ISPs follow, and considerable protection for both the copyright owner and the page owner.
ISPs this, ISPs that. The writer tested TWO ISPs and only ONE responded in an immediately censoring manner. The other provided a detailed questionaire for more information. I'm not suggesting the author doesn't raise a valid freedom of speach concern, but I wouldn't call a "secret shopper" experiement of TWO ISPs statistically useful.
If you're going to do secret investigation hidden camera secret shopper style journalism, you need to select a larger group than TWO ISPs.
This one gang kept wanting me to join cause I'm pretty good with a bo staff.
Bah! Don't try to confuse me with the truth ;)
I disagree with your view on government, but so what. Censorship is censorship. Just because you define it to only apply to government (and so does the law) doesn't mean its not as dangerous. There is no coincidence that megacompanies are buying up media outlets (like MSNBC) to minimize their negative appearances. With all of these megacompanies merging and merging, its going to become pretty difficult to put out the word when they violate the laws that are meant to protect us. I guess we could use the Internet (oops, I just read the article--I guess not). Maybe the only way to combat them is to get a loudspeaker and go to a public square (oops, they don't exist anymore only corporate malls and corporate football stadiums, etc). Hmm... Well I guess since the government didn't define corporate censorship in the 1700s its not dangerous, so I might as well give in. I mean, the important tenents of government never change, right? Right?
This test is called the Pompous Bastard Test (PBT, for short) and it goes something like this:
First, count how many commas, semicolons or colons a writer uses in his sentences. Second, use the chart below to determine the writer's Pompous Bastard Quotient (PBQ):
x <= 1: Dubyass
1 < x <= 3: Normal
3 < x <= 6: Somewhat Cocky
6 < x: Pompous Bastard
Let's take a few examples from the above mentioned text:
"The struggle between Liberty and Authority is the most conspicuous feature in the portions of history with which we are earliest familiar, particularly in that of Greece, Rome, and England."
3 commas. Seems pretty Normal so far.
"THE subject of this Essay is not the so-called Liberty of the Will, so unfortunately opposed to the misnamed doctrine of Philosophical Necessity; but Civil, or Social Liberty: the nature and limits of the power which can be legitimately exercised by society over the individual."
Ooooh, 4 commas (2 commas, 1 semicolon, 1 colon). Getting a little cocky here.
"It is so far from being new, that, in a certain sense, it has divided mankind, almost from the remotest ages, but in the stage of progress into which the more civilized portions of the species have now entered, it presents itself under new conditions, and requires a different and more fundamental treatment."
7 commas. Pompous Bastard. Could be a fluke.
"First, by obtaining a recognition of certain immunities, called political liberties or rights, which it was to be regarded as a breach of duty in the ruler to infringe, and which, if he did infringe, specific resistance, or general rebellion, was held to be justifiable."
8 Freaking Commas. Wow, he really is a Pompous Bastard.
BTW, I'm not commenting on the content of the writer's work, just on his Pompousness; and whew, this guy, whether he knows it or not, is a real, God-damn, Pompous Bastard.
ooops.
+1 Insightful, -1 Troll. What can I say, I'm an Insightful Troll.
When my wife and I started our own message board, a competing board tried to get us to stop perfectly legal activities through extortion; i.e., they threatened to contact our ISP and accuse us of spamming if we didn't do what they wanted.
We didn't do it. They contacted our ISP. I had contacted our ISP in advance with a copy of the extortion email, and it was dealt with appropriately. We never went offline.
Now, that's with a good ISP, that I was in open communication with, and that I warned in advance about the contact, and everything was fine. However, if they had contacted our ISP without first tipping me off to their plans, and had claimed that we were hosting child pornography or infringing someone's copyright, things might have been different.
Chances are it would have taken a couple of days to get it all worked out and back online, but what if the site were politically oriented, and the persons who had it taken down had access to the media?
Imagine this scenario:
1. I post legal, factual information that I feel people should know about before voting for big, powerful mogul;
2. Powerful mogul tells my ISP that I'm doing something illegal (that I'm not doing), and the ISP takes my site down without review;
3. Powerful mogul "leaks" to the news that this site was taken down because of illegal material (let's say child pornography);
4. Most news agencies try to contact me, but a few don't, and run the story as-is;
5. My site is still down, so in the crucial 48 hours after the news story runs, I have no web site to post any kind of rebuttal;
6. Three days later, my site goes back up and my ISP says "it was all nonsense", but it's old news by now. My credibility is shot, and my message is rendered ineffective.
The point of all this, I think, isn't that illegitimate takedown requests can be corrected in a few days, but that illegitimate takedown requests can silence your voice for a crucial period of time that a shrewd, powerful manipulator can leverage for maximum advantage (as in my theoretical example above).
...which is yet another black hole in his methodology. (See the full report for details of the methodology.)
Quote from the article:
The UK ISP took Mill down almost immediately (in huge letters, as a heading, followed by:)
The US ISP followed up on the dubious complaint, made on behalf of the chairman of the non-existent John Stuart Mill Heritage Foundation, with detailed questions. But the UK ISP took the site down almost immediately, effectively censoring legal content without investigation.
In the full report, he details how the US ISP refused to take down the content *unless he explicitly perjured himself in writing*, acknowledging in writing the phrase "under penalty of perjury". Unsurprisingly he was unwilling to do this and terminated the project - so the US ISP left the content there, and did exactly the right thing - a fact he completely skips over in the summary article.
The reason it did do this was that the DCMA provides a framework for such complaints. While ultimately his point in the article is that government regulation may be preferable to private corporate censorship, he doesn't want to explicitly draw the logical conclusion that in this case the DCMA is working exactly as intended, protecting the alleged infringer against specious claims. Rather he decides to reference "anecdotal evidence" that the DCMA has chilling effects.
Now I believe the DCMA to be one of the most wrong-headed laws ever signed, and I agree with this guy's conclusions, but the report reads obviously that he has started with these conclusions and moulded his data to fit them. His methodology is completely out the window. And that's a problem, because anyone can come up with a "research report"; that proves whatever they want, including the bad guys - and from a methodology perspective this guy could get a job with them.
I also on a side note the writter of this story doesn't really under stand the meaning of censorship. Only a government can censor a person, a private company does not have this ability.
Also as a side note you cannot censor published content, you can restrict it, but the litteral word of censor is not possible to do on published content. So the government can never censor published content.
Ok now, from The American Heritage® Dictionary of the English Language, Fourth Edition:
censor: A person authorized to examine books, films, or other material and to remove or suppress what is considered morally, politically, or otherwise objectionable.
censoring: (1) counterintelligence achieved by banning or deleting any information of value to the enemy [syn: censorship, security review]
(2) deleting parts of publications or correspondence or theatrical performances [syn: censorship]
Nothing about only the gov't having the power to censor. Anyone or group can censor. See the Wikipedia for more info.
DMCA covers text.
:)
:)
For example, if you decrypt the following and/or post the plaintext I could in theory use the DMCA against you. No, I won't actually do it. I do wonder how soon someone will in fact post a decryption of it here. And how long before someone posts a one-line Perl script decrypter and makes it open-source.
Zpv Tmbtiepu hfflt offe up hp pvutjef npsf.
Just because it CAN be done, doesn't mean it should!
Are there really people like Bill Gates and Donald Trump out to get you?
;-), and the material is old enough, what you're doing is legal everywhere (as far as we can tell).
Yes. Well, maybe not quite as powerful as those two turkeys, but powerful enough.
An interesting case: I've been involved for some years with a crowd that's putting a lot of old music online. It's not at all unusual for some publisher to send C&D letters demanding that some music from the 1800's or earlier be removed because they own the copyright on it.
I've received a few of these, but my web site is on an academic machine, and the admins who run it are well aware of such problems. They just forward such messages to me, and I deal with them.
For people using commercial ISPs, it's common for the ISP to delete such files without notifying the file's owner.
One funny thing is that we often have documentation of when something was first published. We then send back a reply of the form "That music was published in 1735 in London by Georg F Handel; the file in question is clearly not a copy of your publication and the music is public domain. How are you claiming to own the file's content?" We never hear from the publisher again. Not surprising, actually, as we've told them that we can prove that their claim was fraudulent. Their lawyers then advise them to try some other sucker.
Of course, if you scan in a recently published copy of, say, Beethoven's Moonlight Sonata, you are illegally copying a current publication. But if you enter the music by hand using any of the many music editing packages, the result is not legally a "copy" when the music itself is pubic domain. This is something that tends to be obscured or glossed over in most copyright discussions.
We generally do warn people about putting scanned images of published music online, unless you can prove that your paper copy is out of copyright. But if you've created the files yourself without scanning (with anything but your eyes
Still, ISPs can and do take such things down without notice and without recourse. An outstanding question is whether an ISP can be sued for such invalid actions. You can certainly sue a publisher for making a fraudulent copyright claim, at least if they don't back off when you present them with the evidence.
But it's likely that an ISP is immune to prosecution in many countries. They can censor as they like.
Anyone know different?
This could be used as an argument for government-run ISPs. In the US and many other countries, a government agency couldn't censor users' content without a court order, and you have legal recourse. With private ISPs, you apparently have no legal protection at all.
Those who do study history are doomed to stand helplessly by while everyone else repeats it.
also on a side note the writter of this story doesn't really under stand the meaning of censorship. Only a government can censor a person, a private company does not have this ability.
I've heard this from other people as well, and the issue is that you've got the first amendment confused with the definition of censorship.
From Merriam-Webster:
censorship (noun) 1 a : the institution, system, or practice of censoring b : the actions or practices of censors; especially : censorial control exercised repressively
If I run a chatroom for AOL and I keep the chatroom kiddy-safe by kicking objectionable posters and deleting their remarks, I am censoring the chatroom. The First Amendment to the US Constitution, on the other hand, basically states that the government is usually not permitted to censor individuals (free speech) or publishing interests (free press).
If you wanted to be correct, you could say that the first amendment only prevents the US government from censoring content, and that corporations and people are not limited by "free speech" rights to censor themselves and others who speak or publish through them.
At that point, we get into the interesting topics that this issue raises: is it okay if the government asks nongovernment entities to censor for it? Related question: is it good if nongovernment entities universally censor certain opinions they consider unpalatable (say, antiglobalization)?
But your definition of censorship is in error.
Regards,
Ross
Very true. A better analogy...
A better analogy would be to stop it with all the analogies.
Look at it on its own merits. Something that is true for a similar situation may not be true for another situation even though it appears similar. The concepts here are not hard to understand. No analogy is needed. Is it censorship if an individual can have content he disagrees with removed at will? Simple question. Keep to the point.
I host about 20 domains for clients on my server.
I once received a letter from lawyers asking me to put a client's site down.
I called the client then only replied with an email saying something like "You say it's illegal, but I understand my client's lawyers disagree. I'm only an ISP, not a lawyer, so I will let you sort this out between professionals. I will close the site if and when I receive an official decision from a judge telling me to do so."
I didn't even word it in a more sophisticated way than above. That email took me 5 minutes. Less than the time to take the site down and notify/apologize/explain the matter to the client (who got a Cc of my reply).
Why would any ISP do it differently and just let down his client, to please some anonymous lawyer?
Of course, I would probably have acted differently if the site had been something I consider clearly unacceptable. But it was just about 2 parties accusing each other of being crooks.
Anyway, I never heard of these lawyers again, but heard much later from my client that they actually did go to court, and were dismissed.
If the site is not *very* obviously illegal, why would an ISP act as a judge?
Isn't there a risk when you are not using primary sources?
A modern edition of Handel might include corrections, later transcriptions and other editorial content still under copyright. I don't trust the notion that a "hand-made" copy is perfectly safe.
I can remove content I disagree with at will from my lawn. Is that censorship, or is that my free speech?
common sense: noun
What those who are ignorant of the subject matter think; usually wrong.