Top EU Court Hands Down Judgment On Upload Filters That Is As Clear As Mud

from the it-does-what-for-who-now? dept

We had just written about the great difficulty national governments are having in transposing the EU Copyright Directive into local law. That’s largely because of the badly drafted and contradictory Article 17. It effectively calls for upload filters, which have obvious problems for freedom of expression because of the impossibility of crafting algorithms that encapsulate the subtleties of copyright law. For this reason, the Polish government brought a legal challenge to Article 17 before the EU’s top court, on the grounds that it infringes on the freedom of expression and information guaranteed in Article 11 of the Charter of Fundamental Rights of the EU. The Court of Justice of the European Union (CJEU) has handed down its judgement, dismissing Poland’s action:

the obligation, on online content-sharing service providers, to review, prior to its dissemination to the public, the content that users wish to upload to their platforms, resulting from the specific liability regime established in the Directive, has been accompanied by appropriate safeguards by the EU legislature in order to ensure respect for the right to freedom of expression and information of the users of those services, and a fair balance between that right, on the one hand, and the right to intellectual property, on the other.

That’s regrettable, but on the plus side, the CJEU has made two comments that impose major constraints on the use of upload filters. First it says:

a filtering system which might not distinguish adequately between unlawful content and lawful content, with the result that its introduction could lead to the blocking of lawful communications, would be incompatible with the right to freedom of expression and information and would not respect the fair balance between that right and the right to intellectual property.

But no filtering systems exist that can “distinguish adequately between unlawful content and lawful content”, because algorithms are unable to gauge reliably whether uploaded material is an infringement or a legal use of material for things like parody or criticism. The CJEU further said:

the providers of online content-sharing services cannot be required to prevent the uploading and making available to the public of content which, in order to be found unlawful, would require an independent assessment of the content by them in the light of the information provided by the rightholders and of any exceptions and limitations to copyright.

This seems to say that online platforms can only block material if it is obviously infringing – for example if it is identical to a copyright work. That’s pretty much what the CJEU’s Advocate General suggested last year, when he offered his preliminary opinion on the case, as is usual. Advocate General Henrik Saugmandsgaard Øe wrote:

sharing service providers must only detect and block content that is ‘identical’ or ‘equivalent’ to the protected subject matter identified by the rightholders, that is to say content the unlawfulness of which may be regarded as manifest in the light of the information provided by the rightholders. By contrast, in all ambiguous situations – short extracts from works included in longer content, ‘transformative’ works, etc. – in which, in particular, the application of exceptions and limitations to copyright is reasonably foreseeable, the content concerned should not be the subject of a preventive blocking measure.

Crucially, the CJEU in its judgment emphasized:

Member States must, when transposing Article 17 of the Directive into their national law, take care to act on the basis of an interpretation of that provision which allows a fair balance to be struck between the various fundamental rights protected by the charter of fundamental rights.

In other words, except for the German and Austrian Article 17 implementations, which attempt to provide for ex-ante safeguards, as pointed out by former MEP Felix Reda, this means that many Member States will have to redo their homework in the implementation stage or revisit the legislation they adopted. Unfortunately, the CJEU doesn’t give any guidance on how exactly that “fair balance” can be struck. Which means Article 17’s upload filter provisions remain as clear as mud, and will be the subject of argument and further legal actions for years to come.

Follow me @glynmoody on TwitterDiaspora, or Mastodon.

Originally posted to the Walled Culture blog.

Filed Under: , , , ,

Rate this comment as insightful
Rate this comment as funny
You have rated this comment as insightful
You have rated this comment as funny
Flag this comment as abusive/trolling/spam
You have flagged this comment
The first word has already been claimed
The last word has already been claimed
Insightful Lightbulb icon Funny Laughing icon Abusive/trolling/spam Flag icon Insightful badge Lightbulb icon Funny badge Laughing icon Comments icon

Comments on “Top EU Court Hands Down Judgment On Upload Filters That Is As Clear As Mud”

Subscribe: RSS Leave a comment
51 Comments
This comment has been deemed insightful by the community.
Anonymous Coward says:

for the right to freedom of expression and information of the users of those services, and a fair balance between that right, on the one hand, and the right to intellectual property, on the other.

Those are not alternatives, as user content is also under copyright. It is phrasing like this that makes it look like politicians have bought into the idea that corporate copyrights bought and owned copyrights are much more important than the creators copyright in their original works.

Anonymous Coward says:

Re:

The one silver lining is that punishments and penalties might happen as a result of failing to meet the impossible task. But realistically, we all know that vested interests in copyright have absolutely no problems with taking down legitimate content. Neither will they be held responsible for doing so – or if they are, they’ll simply pass on the consequences to someone else.

This comment has been deemed insightful by the community.
Doug Wheeler says:

How would algorithm know if permission was obtained?

The issue with any of these automated systems is that it is impossible to know if the poster has permission to post the content. It may be a copyright violation for one user, but not another. We’ve already seen the obvious cases, where an automated system blocked the content owner from posting the content, but what if I have an agreement with a copyright holder that allows just me to post certain content? How would any automated system know that I have a contract in my desk drawer or an email in my in-box with permission to share copyrighted information?

Naughty Autie says:

Re:

Another question would be how an automated filter can tell something that doesn’t require copyright permission, such as something falling under an exception or an original performance of a Public Domain composition. Hell, sing some of my parodies into SoundHound and it identifies the original songs, so an automated filter is definitely going to struggle.

PaulT (profile) says:

Re:

This is one of the major problems with trying to hold 3rd parties liable for copyright enforcement. There’s no central database to query to know if something is copyrighted. Even if there was, there’s no way to know individual agreements you weren’t party to that deviate from the standard copyright licence. Even if you get details of such an agreement, that agreement can be revoked at any time without notifying anyone.

Basically, unless any of these schemes involve a return to copyright having to be registered with a central authority, it’s impossible to accurately filter anything – these will always be false positives and false negatives.

terop (profile) says:

Re:

How would algorithm know if permission was obtained?

One easy was would be to build connections to both the authors and the users, and the online platform could watch the contract’s negotiation. For example, sketchfab is doing this currently, when they allow authors to post content and decide its license, and they can see users paying for the content and obtaining licenses to the content. This doesn’t solve the problems for the “first upload”, but instead it controls the secondary market for the works.

PaulT (profile) says:

Re: Re:

“One easy was would be to build connections to both the authors and the users, and the online platform could watch the contract’s negotiation.”

…and if the contract is negotiated offline and/or behind closed doors?

“For example, sketchfab is doing this currently, when they allow authors to post content and decide its license”

Yes, you’re right – a system on a single platform with a centralised database that can be queried does allow things to happen that are impossible when everybody on the planet can created content without any single platform being advised of it in any way.

The only way to achieve this for general copyrighted material is to go back to a system where copyright has to be registered instead of being automatically granted. Otherwise, no platform can possibly know of the correct state of a work. Despite your regular claims that the platform you created to block as much useful content as possible can do this.

“This doesn’t solve the problems for the “first upload””

You know an ideas’s bad when even you have to admit that your solution doesn’t work properly for its claimed purpose.

Anonymous Coward says:

Re: Re: Re:

The only way to achieve this for general copyrighted material is to go back to a system where copyright has to be registered instead of being automatically granted.

And with the number of works being created and published per minute being orders of magnitude greater than works accepted by traditional publishers, how can such a system be built that does not give publishers an advantage? What happens when someone claims a work previously published with the copyright being registered?

Copyright, and copyright registration worked when unregistered works where limited to works not yet accepted for publication, and only the author and publishers had copies of the work, and reintegration was only carried out for the fraction of a percent of works that the publishers accepted for publication.

Kelly Gray says:

Re: Re: Re:2

And with the number of works being created and published per minute being orders of magnitude greater than works accepted by traditional publishers, how can such a system be built that does not give publishers an advantage? What happens when someone claims a work previously published with the copyright being registered?

Probably by using the system that used to be in place. The creator of a work can submit it for registration before getting a publisher involved. Once the work is registered, nobody else can re-register the same work. Until the work is registered, there is no copyright.

As long as the formalities of registration are simple and cheap, the big corporations won’t have a copyright advantage over a creative individual.

Anonymous Coward says:

Re: Re: Re:3

As long as the formalities of registration are simple and cheap, the big corporations won’t have a copyright advantage over a creative individual.

Assuming that a system can be created to keep up with the rate at which works can be created and registered. Note such a system has to be fully automated, as it takes less than a minute for more works to be published on the Internet than are published via a traditional publisher.

Limit copyright registration to works published by a traditional publisher, and you ensure that the Internet is destroyed, because the traditional publishers cannot stand the competition that it enables.

terop (profile) says:

Re: Re: Re:4

Assuming that a system can be created to keep up with the rate at which works can be created and registered.

such registration system doesn’t need to be global or contain all the works on the planet.

It’s enough that platforms that use the products will check some minimum pieces of information like who is the original author of the material, what steps were done to create the material, what kind of marketing effort was done for it — and then reject the works which cannot prove that they’re authorised by the original copyright owner.

Anonymous Coward says:

Re: Re: Re:5

It’s enough that platforms that use the products will check some minimum pieces of information like who is the original author of the material,

Have you mot heard of plagiarism? That step alone needs a database of all works that have been published, and comparison of a new work against that database, as anybody can claim to be the author of any work.

terop (profile) says:

Re: Re: Re:6

That step alone needs a database of all works that have been published

This db isn’t required. Its enough to logically crosscheck consistency of the information that would-be authors are claiming. If their declared authoring process cannot reproduce the work products that they’re publishing, it’s a good reason to reject the submission.

The following checks are necessary: (no db needed)
1) compare creation process to the work products
for checking if all content published can be created
by the declared process
2) check that the declared process is legally valid / filter out processes which directly says they’re pirating the material

Anonymous Coward says:

Re: Re: Re:7

1) compare creation process to the work products
for checking if all content published can be created
by the declared process

What factors will you use to determine that, as Hollywood level productions are within reach of the individual or small group. Note the Individual has available software every bit as good as that available to Hollywood. Also, you are an outlier in your inability to co-operate with other people.

2) check that the declared process is legally valid / filter out processes which directly says they’re pirating the material

A plagiarist is going to declare that only in the alternative world that you inhabit.

terop (profile) says:

Re: Re: Re:8

if hollywood level production is being published, then his creation process just need to declare how he got 2 million budget approved by his accountant.

but current situation is that platforms are not even interested about the creation process since they actually want premium pirated material reappear in their plstform

Anonymous Coward says:

Re: Re: Re:9

his creation process just need to declare how he got 2 million budget approved by his accountant

Of course, Hollywood accounting among other bookkeeping tricks means that any meaningful “budget” is going to be severely skewed depending on how much money gets to be hidden away from the taxman. But you knew that already.

terop (profile) says:

Re: Re: Re:

a centralised database that can be queried does allow things to happen that are impossible when everybody on the planet can created content without any single platform being advised of it in any way.

platform does not need to do business with you, if your dealings cannot be verified properly. Banks do not deal with people whose identity cannot be verified, so why can’t ownership verification be made mandatory.

terop (profile) says:

Re: Re: Re:3

offer actual verifiable evidence that you wrote meshpages.

1) copyright notice in the source code, with git repo which says copyright notice have not been changed since 2012.
2) ludumdare game releases
3) my 100 games list, all using meshpages engine: https://tpgames.org/games100.html
4) my ownership of the meshpage domain name and dns entries
5) copyright notice in my web page
6) my ability to answer deep questions about structure of the source code and related technologies
7) you can dig irc folks from internet who watched me build the system
8) you can ask my brother if he watched me build the system

Verifiable proof is readily available if just you’ll write a platform that can utilize this information. Letting users input this data to the platform is the first step in solving the problem.

Anonymous Coward says:

Re: Re: Re:4

Those are all claims made by you. None of them are independent proof of authorship. You can prove you own the domain, but that is not proof that you created meshpages.

(Knowing the deep structure of a program could be because you studied it, and not knowing it does not mean someone did not write it).

terop (profile) says:

Re: Re: Re:5

You can prove you own the domain, but that is not proof that you created meshpages.

Our govt checked my meshpage.org ownership by making me transfer the site to inside computers that I own. Dreamhost that originally hosted the site cost like 660e/year, and our government doesn’t want to pay for it, so they asked me to transfer the site to computers that I already own. This transfer was successful, and once dns was updated, govt was happy and they can be sure that dreamhost doesn’t slurp another 660e yearly charge from my bank account.

This transfer from dreamhost to my own computers proves that I own the material, given that the transfer was successful.

Anonymous Coward says:

Re: Re: Re:4

government need to step in and provide the support

I get that Finland is a welfare state, but that’s mostly for those who can’t support themselves. You’re clearly not under that category; you’re far too obnoxious.

And let’s say the government did give you the minimum required welfare, that wouldn’t amount to “millions”. Neither would that guarantee people use your software. But keep waiting, or maybe you should get a move on and sue people like your now-arrested friends at Prenda Law did.

terop (profile) says:

Re: Re: Re:5

But keep waiting, or maybe you should get a move on and sue people

Sure, I could do that, but what exactly would be the grounds for suing them? It’s not against the law if they refuse to give their money to me. Which leads us to the real probelem:

you’re far too obnoxious.

This is what they say about gypsies, when you don’t want to give money to them. So when they actually try to do something useful instead of stealing, you just refuse to give the money. Soon they have no other choice than start stealing when the markets cannot provide a living. It’s this general tendency to refuse to purchase a product if the person offering it wasn’t up to your standards. You wouldn’t look at the product itself, but some imaginary generalized information about status of the person in the society.

This is why government need to step in and balance the equation a little. If some group of people are being harrassed alot in the marketplace, government’s support for them needs to be higher. For example, there was news item in local newspapers few weeks ago that a shopkeeper who wears gypsy dress when selling food items, the customers kept calling for police when they saw a gypsy behind the cash register, even though she was hired to sell food to customers.

Basically your refusal to hand over the money is at this kind of problems. Your critique weren’t placed on the product features or quality of the product, but instead your critique was against the persons who tried to sell it to you. This is illegal practise, even though everyone knows it happens regularly.

So while we found good grounds to sue your ass, I’ve decided not to take that route. I will refuse to sue your ass until it is absolutely necessary.

terop (profile) says:

Re: Re: Re:7

That’s a lot of text just to say you won’t sue anyone because you know you’d lose, due to a complete lack of legal basis.

Your biggest problems aren’t always copyright related.

Yes, the government is not obliged to punish people on your behalf because they don’t give you money for a free offering.

Last week there was news items where criminals were agreeing to pay half billion bucks or 250,000 euros for their pirate sites. If they agreed on that, how much criminal money they have on their bank accounts? Don’t you think these kind of criminals deserve to be punished?

Anonymous Coward says:

Re: Re: Re:8

Your biggest problems aren’t always copyright related.

You think Russia invaded Ukraine based on copyright claims. The RIAA and MPAA have on multiple occasions made every outlandish claim they could, declaring that copyright infringement contributed to communism, rapes, murders, and every malady of social unrest. Most problems aren’t copyright-related, but tools like you have consistently made it your end goal for everything to be copyright-related.

If they agreed on that, how much criminal money they have on their bank accounts?

You do know that taking plea deals are a thing, right? It’s not unheard of for people to avoid harsher penalties by pleading to end the trial. As for how much money they have, it helps to look at how your copyright enforcers have handled other high-profile cases. The Pirate Bay? Fined several thousand Euros, which they couldn’t pay up. ISOHunt? The MPAA attempted to milk the owner out of several million dollars, which they admitted in the court filings they knew he didn’t have since Day 1 of filing the case. The simple truth is that running a site for hosting pirated files is not the moneymaker copyright enforcers try to paint it as.

Don’t you think these kind of criminals deserve to be punished?

I’ve seen the way your copyright enforcers go after the innocent. Any arrest or accusation you make has to be taken with a grain of salt, especially coming from someone like you, who supports organ harvesting and rape and murder because you think copyright got infringed.

terop (profile) says:

Re: Re: Re:9

The simple truth is that running a site for hosting pirated files is not the moneymaker copyright enforcers try to paint it as.

This isn’t true. The proof is in the pudding:
https://www.is.fi/digitoday/art-2000008797285.html

This article (once translated) says that some netradio site generated 300,000euros simply because so many people used the site, but they only got 250,000euros damage award for the activty.

Anonymous Coward says:

Re: Re: Re:10

but they only got 250,000euros damage award for the activty

This doesn’t mean a thing. Prosecution in such cases often claim that more money is made by pirate sites than actually exists. The ISOHunt case was such an example. You can claim damages all you want – when it comes to actually retrieving that money, you have a terrible track record.

Anonymous Coward says:

There’s no filter that can tell what is fair use, reviews, parody, public domain versus infringing content but the problem is this law is badly worded and vague and it was designed to help large media corporations like Disney Sony in an age when millions of people are creating music , video content ,audio, podcasts without signing up with TV or media company’s , filters could simply say we won’t block audio content under 2 minutes or video content under 10 minutes to allow for fair use, reviews ,
It’ll be years before this law is really understood or is there any practical way to make it work without blocking free expression or fair use content . When bad laws are made like this it just creates chaos or people ignore it, like most people under 21 in the USA make fake ids in order to buy beer.
It sounds like media corporations maybe could just send a list of films, TV programs music videos with some kind of digital signature code something like YouTube uses for dmca purposes but even if someone built such a system who would pay for it?
would smaller websites have acess to the data or would they just block all user videos just to avoid building a filtering system

Anonymous Coward says:

Yeah, so, as far as the “working” scenario goes, the arguments/suggestions i have seen amount to this: The onus is placed on the online services to make sure their filters work properly. So, someone has to nerd the hardest any human has ever nerded, then everyone has to buy the universal acme omni-filter from them or get dinged for erroneous false positive and false negative blocking. Very probably with (a) filter review board(s) involved.

Joshua Taylor says:

Meanwhile, over here in the US, The copyright office is requiring upload filters on all websites including Deviantart and Pinterest.

Upload filters

  1. Have so many flaws, the independent artist who uploaded their original art can be mistaken as copyrighted by someone else. Putting Indie artists at risk.
  2. Mandatory upload filters is a nail in the coffin for archiving. The Internet Archive will be forced to shut down. They oppose mandatory upload filters
  3. It’s a death to emulating games.
  4. Fan artists are also at risk because of mandatory upload filters.
  5. It’s worse for public domain.

Add Your Comment

Your email address will not be published. Required fields are marked *

Have a Techdirt Account? Sign in now. Want one? Register here

Comment Options:

Make this the or (get credits or sign in to see balance) what's this?

What's this?

Techdirt community members with Techdirt Credits can spotlight a comment as either the "First Word" or "Last Word" on a particular comment thread. Credits can be purchased at the Techdirt Insider Shop »

Follow Techdirt

Techdirt Daily Newsletter

Techdirt Deals
Techdirt Insider Discord
The latest chatter on the Techdirt Insider Discord channel...
Loading...