AI Policy and Regulations in the European Union (2020–2025): A Comprehensive Overview

The European Union has become a global leader in establishing a regulatory framework for artificial intelligence that is both comprehensive and enforceable. Its approach combines a strong emphasis on fundamental rights with strategic investments in infrastructure, research, and innovation. The centrepiece of this framework is the AI Act, the first of its kind in the world, which will take full effect in August 2025. The Act uses a risk-based classification, banning applications considered to be of unacceptable risk, such as social scoring, and imposing strict compliance measures for high-risk systems. This landmark law operates alongside other key legislation, including the Data Act, the Digital Operational Resilience Act, the NIS2 Directive, the European Media Freedom Act, and the revised Product Liability Directive, creating a harmonized regulatory environment across all 27 member states.

Recent Legal Regulations (2020–2025)

The AI Act is the centrepiece of the EU’s legal architecture for AI, setting a new global benchmark for regulating emerging technology. It requires transparency for general-purpose AI models, imposes obligations for high-risk applications, and establishes enforcement mechanisms to ensure compliance. Complementary laws such as the Data Act enhance data sharing and interoperability, while the NIS2 Directive strengthens cybersecurity requirements for critical infrastructure. Other measures, like the European Media Freedom Act, aim to safeguard pluralism and transparency in the information space, ensuring AI technologies do not undermine democratic values.

Government AI Action Plan

The EU’s AI strategy is guided by the Coordinated Plan on Artificial Intelligence, updated in 2021 to align the priorities of member states, EU institutions, and associated countries. This plan is focused on accelerating AI development, facilitating adoption, and ensuring ethical use across all sectors. It is supported by programs such as GenAI4EU, which promotes the application of generative AI in key industries, and InvestAI, an ambitious initiative aiming to mobilize up to €200 billion in funding from public and private sources. Significant investment also comes from Horizon Europe and the Digital Europe Programme, with at least €1 billion per year dedicated to AI research and development.

Intellectual Property & Data Usage

The EU’s intellectual property framework supports innovation while safeguarding creators’ rights. The Unitary Patent System and the Trade Secrets Directive ensure harmonized protections across member states. In the context of AI, the Copyright Directive (2019/790) establishes rules for text and data mining, allowing certain uses for research purposes while preserving the right of authors to opt out of commercial exploitation. Under the AI Act, developers of general-purpose AI models must provide summaries of their training data and comply with copyright obligations before placing systems on the EU market.

AI Outputs & IP Protections

Current EU law grants copyright protection only to works that demonstrate substantial human creativity, meaning that fully machine-generated outputs are excluded. Recent case law has clarified that AI-generated works without human creative contribution are ineligible for protection, though human-guided AI outputs may qualify. The AI Act reinforces this position by requiring transparency in training data and ensuring that intellectual property rules are upheld.

AI Investments & Computing Power

The EU is making significant investments in AI infrastructure to match its regulatory leadership. Through the EuroHPC Joint Undertaking, Europe is building some of the world’s most powerful supercomputers, including LUMI in Finland, LEONARDO in Italy, and the upcoming JUPITER exascale system in Germany. These resources will be available to researchers, startups, and SMEs to train and deploy advanced AI models. The InvestAI initiative will also fund the creation of AI gigafactories equipped with tens of thousands of high-performance chips, aimed at boosting computational capacity across the continent.

Judicial Decisions on AI

The judiciary is playing an increasingly important role in shaping AI governance within the EU. In 2023, the Court of Justice of the European Union ruled that AI-generated credit scoring can qualify as automated decision-making under Article 22 of the GDPR, even when nominal human oversight is involved. This decision underscores the importance of safeguarding individual rights in AI-driven processes. National courts have also begun clarifying the boundaries of AI authorship, reinforcing the principle that human creativity remains central to intellectual property protection.

Conclusion

The European Union’s approach to AI is both comprehensive and forward-looking, combining a robust legal framework with large-scale investments in research, infrastructure, and skills development. As the AI Act and related measures come into force, the EU will not only regulate the technology within its borders but also set a global standard for how AI can be governed responsibly and ethically.

For a detailed analysis of the EU’s AI legislative framework, strategic priorities, and infrastructure initiatives, read the full report:AI Policy and Regulations of the European Union – A Comprehensive Report .