Law on paper does not guarantee law in action. As important as the EU AI Act’s enactment will be its enforcement. Enforcement is critical for the Act’s credibility on the market: if the legislation is not properly enforced, regulated entities will not comply with it, nor will they invest in tools to facilitate compliance.

Building upon our early work in our memo “Trust in Excellence & Excellence in Trust,” and in conjunction with our Reflection Group on General-Purpose AI and Foundation Models, this workstream aims to provide actionable blueprints for the mechanisms in which the EU AI Act will be enforced.

In early 2023, TFS carried out a study to compare institutional models for a European Union-level body responsible for overseeing the implementation and enforcement of the AI Act. This research is presented in our memo: Giving Agency to the AI Act. In this memo, we found that an AI Agency with dedicated staff and legal personhood would on balance be more suitable than an AI Board. We were glad to see that the Parliament’s negotiating position, adopted in June, put forward an AI Office with specifications leaning toward the direction of our findings.

In October 2023, TFS published A Blueprint for the European AI Office, which sought to answer the question: What mechanisms would enable the European AI Office to function effectively, efficiently, coherently and legitimately? We identify mechanisms that would enable the AI Office to satisfy these objectives. We conducted desk research and expert interviews to analyse various design features of relevant institutions and the proposed European AI Office. Based on a review of historical examples and solicitation of expert opinion, we developed recommendations for the AI Office, spanning legal, structural, financial, functional, and behavioural mechanisms.

Related resources