Breaking the Balance: Why Tampering with the EU Copyright Directive Now Could Backfire

Picture this: Europe’s digital copyright rules, painstakingly hammered out after years of debate, are suddenly back on the table. Tech giants, music labels, and AI startups hold their breath. The European Commission is circulating a questionnaire, floating the idea of reopening the 2019 Directive on Copyright in the Digital Single Market (DSM). But behind the scenes, industry insiders warn that disturbing this fragile legal equilibrium could unleash a wave of uncertainty, lawsuits, and market distortions-especially as the AI revolution accelerates.

The High Stakes of Reopening DSM

When the DSM Directive came into force, it aimed to bridge the gap between copyright owners and digital innovators. It introduced exceptions for text and data mining (TDM)-essential for AI research-while preserving rights for creators. Yet, as generative AI booms, questions swirl around whether these exceptions go too far or not far enough. Can AI developers freely scrape music, books, and images for training? Should copyright holders have more say, or is the market for licenses already working?

The heart of the controversy is Article 4, which allows TDM for commercial purposes under certain conditions. Critics say ambiguity over what counts as “lawful access” or a valid “opt-out” has led to confusion, disputes, and even legal action. For instance, can an AI company use YouTube music if the platform’s terms forbid it? What if rights holders post a notice, but no standardized protocol exists? The DSM’s language leaves room for interpretation, but industry practice-like using robots.txt files or explicit online declarations-has largely filled the gaps.

AI, Music, and the Illusion of a Broken Market

Some AI firms push for collective licensing schemes, claiming it’s too difficult to negotiate with every rightsholder. But the music industry rebuffs this, pointing out they’ve already licensed millions of tracks to digital services. The real issue, say critics, is not market failure but reluctance by tech firms to pay for content. Imposing mandatory collective licenses or new exceptions would essentially subsidize big AI at the expense of creators, undermining the value of intellectual property.

Enforcement, Not Overhaul

Instead of rewriting the rules, experts urge the EU to focus on enforcement-especially transparency and traceability requirements in the new AI Act. Without robust records and audit trails, AI training remains a black box, making it impossible for rights holders to know if their works are exploited. Stronger procedural tools, such as shifting the burden of proof in copyright disputes, could make a bigger difference than legislative tinkering.

Conclusion: The Perils of Unsettling the System

The DSM Directive represents a hard-won compromise between innovation and protection. Reopening it now, with the AI and digital content landscape in flux, risks creating more problems than it solves. The call from the trenches is clear: enforce the laws already in place, close enforcement loopholes, and resist the temptation to rewrite the rulebook every time technology moves faster than policy. In the turbulent world of digital rights, stability may be the EU’s most powerful tool.

WIKICROOK

• Text and Data Mining (TDM): Text and Data Mining (TDM) automates the analysis of large datasets or texts to uncover patterns, supporting cybersecurity, AI, and research applications.
• Directive (EU) 2019/790 (DSM Directive): The DSM Directive updates EU copyright law, impacting online platforms, AI, and digital content sharing to support a unified digital single market.
• Opt: Opt is a browser signal that tells websites you don’t want your personal data shared or sold, enhancing your online privacy and control.
• Extended Collective Licensing (ECL): ECL allows organizations to license copyrighted works for all rightsholders, including non-members, simplifying permissions for large-scale or collective uses.
• AI Act: The AI Act is an EU regulation setting rules for safe, ethical use of artificial intelligence, including standards for high-risk systems like deepfakes.