Now that the (greater than possible) remaining of the EU AI Act is offered, and constructing on the evaluation of my now formally revealed new monograph Digital Applied sciences and Public Procurement (OUP 2024), I’ve put collectively my evaluation of its affect for the procurement of AI below EU regulation and uploaded on SSRN the brand new paper: ‘Public Procurement of Synthetic Intelligence: current developments and remaining challenges in EU regulation’. The summary is as follows:
EU Member States are more and more experimenting with Synthetic Intelligence (AI), however the acquisition and deployment of AI by the general public sector is presently largely unregulated. This places public procurement within the awkward place of a regulatory gatekeeper—a task it can not successfully perform. This text gives an summary of current EU developments on the general public procurement of AI. It displays on the slim scope of utility and questionable effectiveness of instruments linked to the EU AI Act, reminiscent of technical requirements or mannequin contractual clauses, and highlights broader challenges in the usage of procurement regulation and follow to control the adoption and use of ‘reliable’ AI by the general public sector. The paper stresses the necessity for another regulatory strategy.
The paper could be freely downloaded: A Sanchez-Graells, ‘Public Procurement of Synthetic Intelligence: current developments and remaining challenges in EU regulation’ (January 25, 2024). To be revealed in LTZ (Authorized Tech Journal) 2/2024: https://ssrn.com/summary=4706400.
As this shall be an space of competition and steady developments, feedback most welcome!