Why Article 17 DSMD should be invalidated? [talk transcript]

This is a slightly edited version of the transcript of my talk given at the conference hosted by the Institute of Political Studies of the Polish Academy of Sciences and the Faculty of Law and Administration of the Adam Mickiewicz University in Poznan in December 2021. The talk is also available as a video-recording here (1:38:00) and is based on my work-in-progress paper available here.

I will split the presentation into three different sections and a conclusion.

First of all, I’d like to talk about the scope of Article 17. Christoph and Justin covered the debate about what are the preventive duties under Article 17, and I think most of the conversation these days is about that. It’s about what preventive duties are implied in Article 17; which means that we are not really discussing the scope of licensing obligations and that these obligations are, perhaps, having some fundamental rights issues as some authors, like Professor Spindler were arguing in literature. We’re only discussing the filtering part. But I just want to highlight this because the Member States can do all sorts of things, and they might not always end up in the same place. So, we are discussing the purpose and scope of preventive duties.


You have heard from Christophe the conversation that the Advocate General, and the Commission, and academics have about what actually article 17 requires. In a nutshell, the law is very vague. We do not know what kind of thresholds or what kind of limits there should be on technology. But we all understand that those limits are important to prevent the overuse of technology in the areas where technology is not up to the task.


For me, in this intervention, I am not concerned with the question of this threshold. But with the question of who should actually determine the threshold?


The Advocate General says that it is for the Member States and the Commission to determine the detailed rules when it comes to the threshold and the scope of the preventive duties.


Now, first of all, I take the issue with this being the Commission’s role. I don’t think it’s up to the Commission to do this. I welcome that the Commission is thinking about this in its guidance, but I don’t think it has any mandate to do this. There’s nothing in article 17 saying that the Commission can specify the threshold. It cannot issue binding acts. The question has to be determined either by the Member States or the EU legislation as interpreted by the Court of Justice. There is no room for the Commission to set up those thresholds.


In light of the recent CJEU case law about the validity of soft law instruments, I think it’s quite clear that if they tried to go too far, someone could easily attack the guidance as just overstepping the mark. So that’s my first issue. I don’t think it’s up to the Commission. The Commission can brainstorm and inspire, but that’s all it can do. It’s clear that it’s then up to the Member States because the EU legislation is so vague on this point.


The Advocate General argues that it is the Member States should set the scope and thresholds. Members States may incorporate content recognition tools if they distinguish what is a manifest infringement; and what is big news, this should be done in the national legislation. I agree with that.


The problem is that the reality is that most of the Member states are saying: “we don’t care, we are not going to set up these parameters in the national legislation”. So far, we have one national legislature that decided to do this even before the AG Opinion was out, but most of the legislatures in the Member States are not doing anything. They are just copy and pasting the Directive.


So, the principle is good, but it’s just entirely disregarded by the Member States. So, is that a good enough outcome? Let’s see. Let’s discuss this at the end.


Second, on procedural safeguards. Because, obviously, one thing is to narrow the scope of preventive duties, but second thing is to introduce safeguards to prevent abuses even within the situations that occur when these preventive duties apply.


And here the Advocate General is really thoughtful on many different safeguards. The Advocate General says that safeguards as such are useful, but he also said that the guards on their own are insufficient to create a fair balance. The reality is that most of the Member States are not really capable of delivering any effective safeguards. Because if you only copy and paste a provision of the Directive saying something should not happen, and you copy and paste a provision saying a provider should have effective redress mechanisms, without actually then introducing some mechanism to supervise this and perhaps a sanction when providers don’t do this, then that’s just not an effective remedy. You just have it on paper, and it doesn’t really do anything for the users. And unfortunately, this is what is happening in most of the Member States. They do anchor the principle that you have a complaint and that it should be effective and speedy, but they don’t do anything in terms of supervision, and transparency, and sanctions to actually make sure that that it happens.


It’s like saying you have rights to equal pay without saying that there are any instruments to enforce that claim and that obligation. I think that’s deeply problematic. And it just shows that it is nice to have safeguards but they can’t be only copy and paste of the Directive’s provisions. Copying and pasting article 17 just doesn’t do the job and won’t be compliant with the national constitutions and the EU charter.


A similar thing is happening with the alternative dispute resolution. Article 17 only alludes to ADR and it leaves all the details to the member states. And basically, most of the member states only say: “Oh, and you can agree on arbitration and mediation”. How is that helping anyone? They could have always agreed, even prior to this directive — it’s nothing. There’s no value in that kind of provision.


Judicial recourse? Some member states might anchor this specifically in their implementations. Again, they’re not helping anyone to be able to enforce this, with the exception of some Member States where you have some collective redress mechanism for this sort of things. And transparency is entirely missing because the directive itself doesn’t require much of it, so the Member State legislatures are not really rushing to do much there.


This brings me to my overall point about supervision. I think the Advocate General proposes that all of these procedural safeguards should be implemented by the Member States. And that the scope should be also specified by the member states and that there should be supervision on all of these aspects on the national level. In fact, he says article 17 obliges those providers, and also the administrative and judicial authorities of the Member States, when supervising the implementation of that article, to consider the collateral effect of filtering measures they implement.


Now, in many member states, there’s no such thing. There is no supervision. Unless you say that only supervision is that maybe some user sometime in the distant future will sue a platform. That’s the only safeguard that you get. There’s no administrative oversight. And in most countries, we don’t even try to create collective redress, we don’t try to create some sort of oversight from the ministries when it comes to the handling of the complaints of users. So again, it sounds nice, but it’s just not there.


This brings me to my main point and a conclusion about this. I think the Advocate General’s Opinion is very thoughtful, and it really goes in the right direction. The problem I have with it is that it assumes that the Member States will do something. This is despite the fact that the Advocate General states, as a principle, that the EU legislature dragging the Member States to the situation of very serious interference with the freedom of expression has responsibility for such situation; it is a responsibility of the EU legislature to introduce safeguards on the EU level. At the same time, Advocate General also says that much of the job has to be done by the Member States. In fact, if you read the Opinion, you get the message that the EU legislature did something, but not enough to make this survive on the national level, even from the perspective of the EU Charter.


I think the serious question we should ask ourselves is in what situation do we end up if the court would follow the Advocate General’s opinion. I think the result is as follows. The result will be that the Member States will implement this legislation and then, on the national level, you will have constitutional complaints about insufficient implementation in these Member States. Most of the Member States will have a problem when it comes to procedural safeguards which means that those constitutional courts, in a dialogue with the Court of Justice, would conclude that the national legislatures didn’t do enough to remedy and to safeguard against abuses.


But the strange effect of this is that none of the blame would come back to the European legislature.


The European legislature, if the Advocate General’s advice is followed, would be seen as doing a sufficient job to pass legislation that is constitutionally compliant, but the Member States have most of the work to work out and they would be seen as constantly underperforming. And in that situation, if this is happening in many member states, one keeps asking why didn’t the EU legislature do a better job and shouldn’t it? And did the EU legislature really do a sufficient job in this case?


And this is where I’m sceptical because particularly when it comes to supervision and when it comes to transparency and supervision of the safeguards that are in the Directive, there’s just nothing specific enough in the Directive language. And same goes for the scope and supervision of the scope of preventive duties.


Now, is it acceptable that EU legislation does that and just implicitly passes the ball to the Member States and they have to figure out in the conversation within its constitutional courts what the problem is? And do we accept that EU legislation can be designed in this way? I think that’s for us to answer.


I think that we should really expect a lot from the EU legislature because otherwise, we have some systemic risks for future legislation. While I’m very sympathetic to the proposal of the Advocate General, I think structurally the problem I see is that the Advocate General places a lot of responsibility on the Member States and less responsibility or limited responsibility on the EU legislature. I think the Union legislature should be more responsible.


And I think this is what this legislation is — it’s flawed. I don’t disagree in terms of what is proposed as a potential reading of article 17. I don’t have qualms with that. I’m totally supportive of that actually. I always said that some form of article 17 would be constitutionally acceptable from where I stand. But in the form that it’s designed at the moment with the safeguards that it has, and hasn’t, I think that we have a problem.


I think it’s actually a deep constitutional question because it will influence how the EU legislature has to do its job in the coming years. And so far, the Court of Justice is rather famous for not properly disciplining the EU legislature — the only exception being the area of the fundamental right to data protection and privacy, where the Court has done some work.


And actually, the data protection area offers an analogue situation to this one, which is a little bit easier to understand because it is a vertical situation. Data retention was scrutinized before the Member States and their constitutional courts for many years. Until after many years it finally arrived in Luxembourg and Luxembourg said, well actually, the whole directive is invalid.
I think here we have an inverse of that scenario. We have first the Court of Justice looking at the validity of a directive. And then the Member States, their constitutional courts, figuring out whether the Member State legislatures did enough. I think this is useful to some extent, but it offers some risks, particularly that the Court of Justice says too early that it was enough what the EU legislature did; because I don’t think it is enough.