Position Paper on European AI Regulation
18.12.2025A new policy paper examines the EU AI Act and highlights key risks in its implementation. It focuses on technical standardisation and conformity assessments, warning against excessive reliance on private governance and weakened democratic oversight.
Adopted in 2024, the European Artificial Intelligence Act represents a milestone in AI regulation. In a newly published position paper, the authors welcome its ambition to acknowledge trustworthy and accountable AI, while critically analyzing two core implementation mechanisms: technical standardisation and conformity assessments.
The paper argues that the current design of these mechanisms risks shifting regulatory power towards private actors. Heavy reliance on harmonised standards and predominantly internal conformity assessments may limit democratic oversight, transparency, and effective protection of fundamental rights within the EU’s AI governance framework. To address these risks, the authors call for independent third-party audits for high-risk AI systems, arguing that external oversight is necessary to ensure effective supervision, transparency, and credible enforcement beyond provider-led self-assessment.
The full position statement is now available for download as Policy Paper #17