Standardizing AI - The Case of the European Commission's Proposal for an Artificial Intelligence Act

Martin Ebers
DOI: https://doi.org/10.2139/ssrn.3900378
2021-01-01
SSRN Electronic Journal
Abstract:On April 21, 2021, the European Commission presented its long-awaited proposal for a Regulation “laying down harmonized rules on Artificial Intelligence”, the so-called “Artificial Intelligence Act” (hereinafter: AIA). Most of the provisions of the AIA deal with high-risk systems, setting out obligations on providers, users and other participants across the AI value chain, establishing in particular, conformity assessment procedures to be followed for each type of high-risk AI system.At the heart of the proposal is the idea of co-regulation through standardization based on the New Legislative Framework (NLF). According to Recital (61) AIA, “[s]tandardization should play a key role to provide technical solutions to providers to ensure compliance with this Regulation”. Accordingly, this chapter provides a critical analysis of the proposal, discussing in particular, how the envisaged system of co-regulation, standardization and certification could contribute to European governance of AI and address the manifold ethical and legal concerns of (high-risk) AI systems.
What problem does this paper attempt to address?