Brussels Revisits Copyright – Part 3: Why an Optional Research Exception Is No Exception at All
This is the third and final installment in a series analysing the European Commission’s current call for evidence on EU copyright rules. Part 1 examined live content piracy and the fundamental rights implications of enforcement without safeguards. Part 2 examined the push for mandatory AI training licences and why the mechanism would harm the creators it claims to protect. This post focuses on the research exception: a provision that has been optional for twenty-five years and whose costs are now comprehensively documented.
The Commission’s call for evidence to support its review of the Directive on Copyright in the Digital Single Market (DCDSM) and to lay the foundations for a targeted legislative proposal aimed at strengthening copyright in light of AI and other market developments, which closes on 25 June, explicitly invites views on an updated research exception and a potential EU-level secondary publication right.
Two major studies commissioned by the Commission have now established the evidence base. The reform case from the research and library community is specific, technically grounded, and consistent across multiple consultation rounds. What policymakers need to determine is whether they treat twenty-five years of optionality as a design failure to be corrected or as a settled arrangement to be preserved.
What optional actually produces
An optional exception is not an exception. It is a suggestion. Optional since 2001, the research exception under Article 5(3)(a) of the InfoSoc Directive has been exercised in twenty-seven different ways over twenty-five years, producing twenty-seven different legal environments for cross-border research in the EU. A use that is clearly permitted in one Member State is uncertain in a second and unprotected in a third.
The practical response to that complexity is caution. Institutions decline uses the law in fact permits, because the cost of being wrong outweighs the cost of not trying. That chilling effect operates across the ordinary working conditions of cross-border research, and it has a physical dimension: without a cross-border digital exception for document supply, researchers who cannot access a work through their national digital infrastructure must travel physically to a library that holds it. The inter-library loan network that functions across borders in the analogue world has no functional digital equivalent in EU law.
The exception also contains a structural flaw that exacerbates the fragmentation. Most national implementations draw an artificial distinction between commercial and non-commercial research. Public-private partnerships, which are central to EU research funding policy and built into the Horizon programme’s architecture, routinely cross that line. Under most Member State implementations, the commercial element of a partnership can void the exception for the collaboration as a whole. As a result, the research that benefits most from EU co-funding, and which could potentially contribute most to the EU’s competitiveness agenda, is systematically the research least able to rely on the exception intended to cover it.
The secondary publication gap
Seven EU Member States have introduced secondary publication rights (SPRs), which allow researchers to deposit published work in open repositories, in some cases after a defined embargo period. Those rights do not, however, operate across borders: a researcher in a Member State without a SPR cannot exercise the right by depositing in a Member State that has one.
EC-commissioned research published in 2024 and 2026 found overwhelming support for a harmonised mandatory right in both the research community and among institutional publishers. Publisher support is the more telling finding: it suggests that even those who benefit commercially from subscription models recognise the current patchwork is not a stable long-term arrangement. The case for harmonisation is not only one of open access principle: the same research indicates that a mandatory right could facilitate knowledge valorisation from public institutions and open conditions for innovation in publishing markets, a framing that addresses a different set of objections than the access argument alone.
Two conditions determine whether the an SPR functions in practice or exists only on paper. Contracts must not be permitted to override it: publishers routinely include subscription terms that restrict repository deposits, and individuals and institutions all too often feel obliged to sign away the right to share their work openly in order to get published in so-called ‘high-impact’ paywalled journals. And technological protection measures must not be permitted to block access to lawfully deposited material. Which brings us to the barrier that sits beneath all of the above.
The Technological Protection Measure problem
Technological protection measures (TPM) are the mechanism through which publishers convert advisory copyright restrictions into technical facts. Where an exception exists in law and a TPM blocks its exercise, the researcher faces a choice between a legal right and a technical lock. In most cases, the lock wins.
The Knowledge Rights 21 survey published in December 2024 documents this pattern: researchers regularly encounter TPM barriers to uses explicitly permitted under their national exception, and institutions have no clear legal route to remove those barriers without initiating litigation disproportionate to the individual use at stake. A remedy that requires litigation to activate is not a remedy in any operational sense.
What an enforceable TPM remedy looks like is not a theoretical question. Slovenian law requires that TPM blocks over legally permitted uses be lifted within 72 hours. That specificity is the difference between a remedy and a principle. Updating EU copyright law to strengthen the mandatory research exception without building in operationally specific safeguards against TPMs would reproduce, at EU scale, the gap between entitlement and access that the Knowledge Rights 21 survey documents.
What the Commission should now decide
The call for evidence asked whether the existing research exception framework is fit for purpose. The evidence says no, and the reform asks are specific. Five conclusions follow.
- Make the research exception mandatory: Article 5(3)(a) of the InfoSoc Directive must be made mandatory. An optional exception producing twenty-seven legal environments is incompatible with the European Research Area. This is the baseline from which every other reform measure follows.
- Remove the artificial commercial/non-commercial distinction: Contemporary research does not respect that line, and EU research funding policy is built on partnerships that cross it. Retaining the distinction preserves the chilling effect that makes the exception less useful than it appears on paper, hurting Europe’s competitivness agenda
- Introduce a harmonised mandatory secondary publication right: A zero-embargo right for taxpayer-funded research, under a genuinely open licence, must apply across all Member States. Contracts and TPMs must not be permitted to override it. Seven Member States have already demonstrated that the right is legislatively achievable. What is missing is the cross-border harmonisation.
- Build enforceable TPM remedies into EU legislation: Removal deadlines, institutional circumvention rights, and penalties for repeat offenders must appear in EU law, not in accompanying guidance. Without operational specificity, the TPM remedy will remain as advisory as the exception it is supposed to protect.
- Protect the Article 3 TDM exception from the opt-out process: Article 3 of the DCDSM provides a TDM exception for research organisations acting for scientific research purposes. Any expansion of the commercial opt-out mechanism that narrows Article 3’s scope, or any transparency requirement that functions in practice as a licensing mandate for research TDM, would hollow out the research exception at exactly the moment this assessment is supposed to be strengthening it. The Commission must treat TDM and the research exception raised under the CDSM review as connected questions, not parallel ones.
The available window
Across this series, a pattern has emerged. In Part 1, the urgency of a live event was used to build an enforcement architecture without safeguards, and the Commission documented the failure while Italy compounded it. In Part 2, the claim that creators needed protection was used to advance a mechanism the evidence shows would not protect them. In each case, the Commission has received the evidence it needs and faces a choice about whether to act on it. This post is no different.
The research exception has been optional since 2001. In that time, the EU has watched the United States, Japan, Singapore, Canada, South Korea, Switzerland, and Taiwan settle on clearer legal frameworks for the research uses the EU exception leaves uncertain. The EC-commissioned evidence on what this costs paints a clear picture. The window to correct the 2001 design choice is now. It will not stay open indefinitely.
Written by Caroline De Cock, LL.M., Head of Research
