The US has fair use. Japan and Singapore’s broad, innovation-friendly copyright exceptions privilege AI innovation. In contrast, Europe restricts access to data — with grave risks to its economic future.
The continent’s Copyright Directive grants copyright exceptions for Text and Data Mining but allows rightsholders to exercise the right to opt out. More than half of news publishers already block the main AI web crawlers by using shared standards for robot exclusion.
The consequences of this “data scarcity” are already visible in the widening investment gap between the EU and its global competitors. In 2024, the American private sector invested over €101 billion in AI —roughly 73% of the global total. Europe managed a mere €18 billion. When it comes to Generative AI specifically, the disparity is even more staggering: €27 billion in the US versus just €1.3 billion in the EU.
This isn’t just about missing out on the next ChatGPT. AI is a general-purpose technology that is rapidly transforming the backbone of the European economy. From automotive giants using machine learning to optimize assembly lines to pharmaceutical firms like AstraZeneca accelerating drug discovery, the need for vast, diverse datasets is universal.
The economic stakes are massive. Europe’s creative industries, including art, literature, film, music, theater, and television, are significant, generating €202 billion in value added. The sector argues that unchecked AI data mining represents a threat to their continued prosperity and their contribution to European culture. They want the freedom to opt out of data mining.
But the sectors that need AI to prosper are much larger than the important cultural sector. The combined value of data-intensive sectors such as ICT, finance, automotive, pharmaceuticals, and chemicals is more than nine times larger. To offset just a 1% reduction in value added across these sectors due to poor AI adoption, the creative industry would need to grow by more than 9%. By over-regulating data access to protect one sector, the EU is effectively imposing a tax on its entire industrial future.
Perhaps the most damaging aspect of the EU’s restrictive approach is its impact on small and medium-sized enterprises. Large incumbents can absorb the compliance costs of the GDPR, the AI Act, and the Digital Markets Act. They have in-house legal teams to navigate fragmented “opt-out” landscapes where millions of individual rightsholders might reserve their data.
European start-ups such as Mistral AI or PhariaAI don’t have that luxury. They rely on the open market and the Text and Data Mining exception to acquire training data. If the exception is narrowed or rendered unworkable by expansive opt-out mechanisms, the “law of the jungle” will prevail. Only the largest firms will be able to afford the licensing fees and legal fees, entrenching market power and stifling the very competition EU regulators claim to champion.
To reverse this stagnation, Europe must stop treating data as a scarce concession and start treating it as a strategic asset. This requires a shift in both policy and mindset:
- Protect the Text and Data Mining Exception: The EU should recognize the exception as a core component of its competitiveness strategy. Any proposal to narrow it should be subject to a mandatory economic impact assessment.
- Contain the Opt-Out Mechanism: While rightsholders should have a say, the EU must resist making opt-outs the default or lowering the technical threshold for them. If the process becomes too fragmented, data access becomes impossible for everyone but the biggest players.
- Stabilize the Rulebook: European firms are drowning in a “dense scaffolding” of regulations. Before layering on new copyright restrictions, the EU should allow recently adopted laws like the AI Act to reach maturity.
- Reframe the Problem: Displacement of human labor by AI is a legitimate concern, but it is a labor and market competition issue, not a copyright one. The solution lies in market adaptation and social safety nets, not in breaking the tools of the future.
Europe is not acting in a vacuum. Jurisdictions like Singapore and Japan have already moved toward “innovation-first” frameworks. Singapore’s “computational data analysis” exception permits the use of copyrighted works for data analysis in both commercial and non-commercial contexts and explicitly prevents contracts from overriding these rights. Japan’s Copyright Act allows for “information analysis” regardless of the purpose, provided the expressive content isn’t being “enjoyed” in the traditional sense.
For Europe to remain an industrial and technological power, it must ensure that data remains an abundant, accessible resource. The “copyright trap” is set; it’s time for EU policymakers to step around it. If European policymakers continue to treat copyright law as a defensive shield for the creative sector rather than a strategic lever for the broader economy, they risk turning the continent into a technological museum.
This article is adapted from a longer paper, The Copyright Trap, published by the European Centre for International Political Economy (ECIPE).
Oscar Guinea is a Director at ECIPE. He specializes in international trade, digital markets, and industrial policy. A regular commentator on EU affairs, Oscar writes a column for El País covering trade and economic policy.
Bandwidth is CEPA’s online journal dedicated to advancing transatlantic cooperation on tech policy. All opinions expressed on Bandwidth are those of the author alone and may not represent those of the institutions they represent or the Center for European Policy Analysis. CEPA maintains a strict intellectual independence policy across all its projects and publications.
Tech 2030
A Roadmap for Europe-US Tech Cooperation
