The General Purpose AI (GPAI) Code of Conduct is set to become a key framework for AI model providers, outlining how they comply with the EU AI Act throughout the entire lifecycle of their models. However, concerns have emerged that the current draft of the Code imposes excessive regulatory burdens, potentially stifling AI innovation in Europe, particularly for startups, SMEs, and regional AI actors.
SEEDIG, alongside twelve other organizations, has co-signed an open letter led by the AI Chamber, urging policymakers to revise the second draft of the GPAI Code of Conduct before its finalization in April 2025. The letter, addressed to the European Commission, highlights key regulatory issues and calls for a more balanced approach to AI governance that fosters both safety and innovation.
Why SEEDIG Supports the Open Letter
At SEEDIG 9 annual meeting, the importance of ensuring inclusive and practical AI governance was a recurring theme. As a leading internet governance platform connecting stakeholders across South Eastern Europe (SEE), SEEDIG strongly advocates for open, multistakeholder dialogue on AI policy that reflects the needs of all affected actors, including emerging tech companies, SMEs, and regional digital innovators.
“AI regulation should ensure safety and accountability while also encouraging technological progress. The current draft of the GPAI Code introduces complex compliance measures that could disadvantage European AI companies in the global market. SEEDIG stands with its partners in calling for a more practical and innovation-friendly approach,” – Olga Kyryliuk, SEEDIG Chair.
Key Concerns with the GPAI Code Draft
- The current draft introduces additional obligations beyond the AI Act, increasing costs and compliance burdens on AI model providers. Startups and SMEs, in particular, may struggle to meet these demands, hindering AI development in South Eastern Europe and beyond.
- The draft demands extensive reporting on AI model parameters, training data, and computational resources, raising intellectual property and security concerns. Such measures risk undermining European AI companies’ competitiveness on a global scale.
- The draft is built on AI standards that are yet to be defined, creating uncertainty about actual compliance requirements. This ambiguity makes it difficult for AI developers to plan and innovate confidently.
- The draft introduces new copyright compliance obligations that go beyond existing EU copyright law, creating legal uncertainty for AI developers working with third-party datasets.
- While the draft aims to ensure responsible AI governance, it must also support AI growth within Europe. The EU risks falling behind the U.S. and China in AI development if regulations become too rigid and impractical for companies to implement.
Call to Action: Join the Open Letter
SEEDIG encourages regional stakeholders – from policymakers to civil society and the private sector – to engage in discussions on AI regulation and ensure that the SEE region’s voice is heard in shaping AI policies.
The joint letter remains open for additional signatories who support a balanced, innovation-friendly approach to the GPAI Code of Conduct. To learn more and sign the letter, click here.
This blog is based on content originally published by AI Chamber. Read their full article here.
A more detailed explanation of the concerns surrounding the Second Draft of the AIA GPAI Code of Practice can be found in a previous SEEDIG op-ed here.