Simona Demková discusses the EU’s human-centred approach to regulating artificial intelligence
On 27 and 28 April, Simona Demková participated as a panelist at the conference 'A "Brussels Effect" for EU Digital Governance? Legal, Political and Economic Considerations' organised jointly by the University of Lausanne, Ecole Polytechnique Fédérale de Lausanne (EPFL), the University of Groningen, LUISS, Université Libre de Bruxelles and Wallonie-Bruxelles International. The conference convened with the keynote from the author of the ‘Brussels effect’ phenomenon herself, Professor Anu Bradford from Columbia Law School.
Simona joined the panel on ‘Algorithms and AI’ to discuss the EU’s ‘rights-based’ and ‘human-centred’ approach to regulating AI. By critically examining the genealogy of the approach, her presentation suggested that the EU’s currently pending Proposal for AI Act falls short of ensuring full respect for fundamental rights. Instead, the Act’s product safety nature combined with its risk-based approach and the proposed market surveillance enforcement scheme point to a greater convergence with the ‘human-centred’ vision of the technology industry. The latter, reflecting advances in computer science, essentially strives at warranting effective human control over AI uses, not necessarily greater protection of fundamental rights and human values. This discrepancy will have significant implications on the formulation of safeguards within the final version of the AI Act and, more broadly, on the EU’s ability to meet its ambition of creating the global standard for AI.
Simona will continue to examine these dynamics together with Daniel Mândrescu in the context of a new project on ‘The EU’s Human-Centred Digital Transformation,’ supported by the Leiden University Starter Grant. To contribute to the EU’s successful human-centred digital transformation, this project aims at creating a coherent conceptual framework for the realisation of the human-centred requirements and their enforcement under the EU digital acquis. To that end, the project aims at:
- recalibrating and coordinating the related yet different regulatory objectives of the EU digital acquis;
- exploring how these objectives are currently pursued and what the human-centred regulatory approach will demand in practice from the various actors involved: EU/Member States authorities; technology providers and/or business practices to which the instruments apply;
- identifying the potential frictions between the various objectives of EU digital acquis and the technical design of the AI technology to which it applies.
The project aspires to convene a multi-stakeholder discussion to gather insights into the practical implementation of the ‘human-centred’ safeguards. For more information about possible collaboration, please contact s.demkova@law.leidenuniv.nl.