Two Megaphones and an Empty Room: The Need to Write the Copyright Reform for the Empty Seats
The EU copyright review is being argued by two blocs of private interests. The party that will live with the result, the public, was never really in the room. That was a choice by those who drafted the consultation. That is not a good thing.
Two private interests are shouting at each other across the European Commission’s copyright review, and the room they are shouting in is empty. On one side stand rightholders. On the other stands the artificial intelligence (AI) industry. Both are loud, both are organised, and both are arguing about the same thing: who captures the value of creative and informational works. The Commission’s call for evidence drew just over 430 submissions, and the coverage has settled into a familiar two-sided story. That story is wrong in the way that matters most. It leaves out the only party with nothing to sell but much to loose.
Two private interests, little room for the public interest
Strip the consultation down and the real divide is not rightholders versus AI. It is private interest against public interest. The rightholder camp wants control, remuneration and tighter rules. The AI camp wants room to train and less friction. Each wants a slice of the same pie, and they disagree mainly on the size of the slice. Set against both is a single constituency that is not asking for a payment at all. It wants access to knowledge, the preservation of the record, and enforcement that does not break the open web. That constituency is the public interest, and it appears in this consultation in two forms, one quiet and one silent.
The quiet form is the custodians. Libraries, archives and researchers, the institutions that keep the public record and make publicly funded knowledge usable, filed in force and arrived with the most concrete proposals in the whole exercise. Ninety-seven submissions named the fragmentation of the rules across member states as the core problem. Thirty-nine backed a binding Secondary Publication Right (SPR) to free publicly funded research from the contracts researchers are required to sign. Access, they argue, is a right rather than a privilege. They are in the room. They are simply not being heard over the two megaphones.
The most concrete evidence in the entire consultation comes from one of these custodians. The National Library of Sweden, which administers the Bibsam consortium of nearly a hundred research institutions, did not argue from principle. It described an operation. Where the law grants libraries and researchers the right to mine text and data, publishers refuse the application programming interface (API) access that would make the right usable, or supply technical solutions that do not work, forcing the consortium into extensive additional negotiation to secure what the exception already guarantees.
The right exists on paper. The contract decides what happens in practice.
This is what it looks like when a public-interest entitlement is granted by the legislature and then quietly switched off by private agreement. It is the single most citable piece of operational proof the public interest brought to the consultation, and it explains why the custodians keep asking for the same thing: exceptions that no contract can override.
The silent form is the public itself, and here the numbers are stark. Of 432 submissions, individuals and consumer organisations together account for 41, which is 9.5 percent. Across the entire European Union, exactly two consumer organisations filed. The people who will live with whatever the reform decides, the readers, the users, the citizens, barely registered. This is the empty room.
The room was built empty
It would be comfortable to call that a failure of public engagement. It was not a failure. The consultation was built in a way that was never going to reach the public, and the design choices that produced the empty room were not neutral.
Start with the instrument. This was a call for evidence, the European Union’s lightweight feedback format, and not a full public consultation with a structured, translated questionnaire. The Commission has said plainly that the genuinely public-facing consultation will come later, attached to the impact assessment. So the first pass, the one that gathers the evidence that frames everything after it, was pitched at the people who already hold evidence: industry and institutions.
Then the language. The document does not ask a question an ordinary person can answer. It asks for views on the Commission’s understanding of a problem described in the vocabulary of Article 30 reviews, single equitable remuneration, material reciprocity, the RAAP judgment and rights reservation mechanisms. A person who loses access to a work, or whose unrelated website is blocked by mistake, does not hold a considered position on Article 8(2) of Directive 2006/115/EC. The language register itself is a filter.
Then the guest list. The call for evidence names its target audience, and it reads as a roll call of organised interests: rightholders, AI providers, data aggregators, online platforms, video-on-demand services, collective management organisations, sports and live event organisers, research bodies, heritage institutions, national authorities. Consumers and users appear at the very end of that list, and only as organisations. The individual citizen is not named as an audience at all.
The result matched the design. An institutional instrument, written in institutional language, addressed to a list of institutions, produced an institutional consultation, four fifths of the feedback received in English, with Brussels-based federations the single largest bloc. This was not a public that failed to turn up. It was a public that was never addressed.
The Commission knows how to reach people when it wants to. It runs full public consultations, in plain language, in every official tongue, when a file calls for them. Choosing the narrower instrument for a reform whose costs fall hardest on the public is not a neutral act of administrative sequencing. It lets the insiders set the terms, and pour the evidence base, before the public is ever handed a question it could answer. By the time the promised consultation arrives, the framing will already have hardened.
Absence is not indifference
The public defends what it can see. When a harm is concrete and personal, people mobilise. A campaign against bought video games being switched off after purchase gathered around 1.2 million signatures, and the handful of game-related submissions in this consultation came almost entirely from private citizens. People will fight for a thing they can watch being taken away from them. The trouble with copyright enforcement is that its sharpest harms stay invisible until the moment they land, and by then no consultation is open.
The clearest case is live-sports piracy, the enforcement centrepiece of the whole file. Sports leagues and broadcasters filed in volume for binding, near-instant blocking of pirated live streams. The case against piracy is real. The danger sits in the blunt instrument. As the internet service providers (ISPs) themselves told the consultation, a single internet address can host thousands of unrelated services, so an order aimed at a pirate stream routinely takes down banking apps, charity websites and developers’ tools as collateral damage. An empirical submission from an intellectual property researcher made the same point and supplied the remedy: enforce at the source and follow the money, avoid address-level blocking, keep any block narrow and limited to the duration of the event, and require a proportionality check before it is imposed rather than after.
Read the list of who raised this collateral harm and the empty room comes into focus. It was named by ISPs, by challenger search engines, and by a couple of organisations representing the public interest. It was not named by the users who would lose their banking app for the length of a football match, because those users were never in the room. The harm that lands on the public is argued entirely by proxy, in a technical register, on behalf of people who were not invited and cannot see the order coming. The more technical the measure, the more total the silence.
Write the reform for the empty seats
This is where information labs starts: to make technical policy legible enough that the people it affects can see what is at stake. The reform now being drafted should be written for the public interest, and the institutions speaking for it in this consultation have already supplied the substance.
No contract or technological protection measure should be permitted to override a statutory exception. This is the demand repeated most insistently across the public-interest filings, and it is repeated because the evidence, Bibsam included, shows exactly how the override happens in practice. An exception that exists on paper but cannot be exercised is not an exception at all.
The framework built in Articles 3 and 4 should be defended, not replaced, and Article 4 should be stopped from swallowing Article 3. The Commission does not need a new licensing regime for text and data mining. It needs to confirm, in guidance, that opt-outs designed for commercial data exploitation under Article 4 cannot be applied to the unconditional research and preservation protection the legislator built into Article 3.
The Secondary Publication Right should be legislated, with zero embargo. Publicly funded research should be available the moment it is published, under a truly open licence, and no publisher contract should be allowed to delay that.
Live-event blocking should be walled off from the rest of the internet. The case against live piracy is real, and so is the documented pattern of blocking orders aimed at a football stream taking down unrelated banking apps, charities and infrastructure providers. Any legislative tool built for the urgency of a live broadcast needs a structural prohibition stopping it from reaching general-purpose, non-commercial websites, together with proportionality, prior judicial scrutiny and a fast unblocking mechanism.
The process owes the public a debt as well. The consultation still to come must reach people in plain terms, in their own languages, before the legislative choices are locked, and not as a courtesy after the evidence has already been poured. A consultation that only specialists can answer is not a neutral consequence of a technical subject. It is a decision about who counts at the most critical stage of decision-making.
Two megaphones are switched on. The room is in the dark. The reform is being drafted now, and the only question that matters is whether anyone writes it for the empty seats.
Written by Caroline De Cock, LL.M., Head of Research
