📢THE EU AI ACT: LEGISLATIVE DEVELOPMENTS AND STRATEGIC IMPLEMENTATION INITIATIVES

As the European Union moves forward with implementing the AI Act, the European Commission is unveiling broader strategies to establish Europe as a global leader in trustworthy AI development and deployment. At the heart of this momentum is the newly announced AI Continent Action Plan, a comprehensive roadmap to leverage EU strengths and ensure Europe’s digital sovereignty.

Strategic Vision: The AI Continent Action Plan

The European Commission’s Action Plan is structured around five key pillars:

1. Developing a large-scale data and computing infrastructure

2. Enhancing access to high-quality datasets

3. Fostering the development and uptake of algorithms in strategic sectors

4. Strengthening AI skills and talent pipelines

5. Simplifying regulatory pathways

The AI Act underpins this vision, aiming to create a harmonised regulatory environment that facilitates cross-border AI deployment while ensuring safety, respect for fundamental rights, and market access uniformity across Member States. Recognising that regulatory complexity can be an obstacle, the Commission has announced an AI Act Service Desk, designed to guide providers—particularly SMEs—through the regulatory framework.

Open Source Debate: Preserving Transparency and Competition

Recent developments also reveal growing concern over the definition of “open source” in the Act's implementation phase. A coalition of 30 MEPs has called on the Commission to adopt a strict and meaningful definition that aligns with the Open Source Initiative’s standards. The letter directly challenges Meta’s licensing of its Llama models, arguing that restrictive licensing terms render such AI models incompatible with a truly open-source framework. The broader concern is that a diluted open-source definition may jeopardise citizens’ rights and impede European innovation.

Open Source Debate: Preserving Transparency and Competition

Recent developments also reveal growing concern over the definition of “open source” in the Act's implementation phase. A coalition of 30 MEPs has called on the Commission to adopt a strict and meaningful definition that aligns with the Open Source Initiative’s standards. The letter directly challenges Meta’s licensing of its Llama models, arguing that restrictive licensing terms render such AI models incompatible with a truly open-source framework. The broader concern is that a diluted open-source definition may jeopardise citizens’ rights and impede European innovation.

Building Literacy: Article 4 and the AI Repository

The AI Office is actively supporting the implementation of Article 4 of the AI Act, which addresses AI literacy and awareness obligations. A new living repository has been launched to collate literacy practices from organisations across Europe. While inclusion in this repository does not constitute regulatory compliance, it offers valuable insights into best practices and transparency initiatives. A public survey is now open to organisations seeking to contribute to the evolving landscape of AI literacy.

Competitiveness vs Sovereignty: A Delicate Balance

Prominent voices in the policy space, including Marietje Schaake International Policy Director at Stanford University’s Cyber Policy Center and Max von Thun Director of Europe and Transatlantic Partnerships at the Open Markets Institute, caution against an overemphasis on competitiveness as a guiding principle of EU tech policy. Their concern is that focusing too narrowly on deregulation could entrench the dominance of US tech giants, undermining the very sovereignty and independence the EU seeks to secure. Legal experts and commentators are increasingly warning that meaningful enforcement of digital regulations—including the AI Act—is essential not just for consumer protection, but for creating an environment where European alternatives can flourish.

Regulation and Investment: Two Sides of the Same Coin

As the Code of Practice (CoP) for the AI Act nears completion, the EU’s approach balances regulatory obligations with strategic support. While the most stringent requirements are targeted at large general-purpose AI model providers, the broader ecosystem—particularly SMEs—stands to benefit from greater legal certainty and reduced liability risks. Experts argue that deregulation would not solve Europe’s tech deficit, and instead, upstream regulation could provide the clarity needed for downstream innovation.

Human Oversight: A Work in Progress

Finally, attention is turning to the human oversight provisions in Article 14 of the AI Act. While the article calls for effective human control over high-risk systems through both design and organisational processes, ambiguities in terminology and guidance pose real implementation challenges. A recent report by the Knowledge Centre Data & Society calls for sector-specific standards, enhanced user training, and greater awareness among AI providers and deployers. Without these, the risk is that compliance will be superficial and trust in AI systems undermined.

Conclusion

The months ahead are critical for stakeholders involved in AI innovation, deployment, and compliance across the EU. As the Commission rolls out support mechanisms and finalises practical guidance, the legal community must stay actively engaged. From defining open-source frameworks to clarifying oversight responsibilities, the details of implementation will determine whether the AI Act delivers on its promise of a secure, rights-based, and innovation-friendly AI ecosystem in Europe.

Our firm is closely monitoring the regulatory evolution and stands ready to advise clients navigating the complex but promising terrain of AI compliance in the EU.

For more information on how the AI Act and related regulations may affect your business, or to receive tailored legal guidance on regulatory compliance, please contact info@pelaghiaslaw.com. Our team is available to support your organisation in understanding, implementing, and leveraging the legal frameworks that shape the future of AI in Europe.