Publications

The use of behavioural insights (BI) in public policy has grown over the last decade, with the largest increase of new behavioural teams emerging in the last five years. More and more governments are turning to behavioural science – a multidisciplinary approach to policy making encompassing lessons from psychology, cognitive science, neuroscience, anthropology, economics and more. There are a wide variety of frameworks and resources currently available, such as the OECD BASIC framework, designed with the purpose of helping BI practitioners and government officials infusing behavioural science throughout the policy cycle.

The European Union (EU) is engaged in a complex digital and ecological transition. The policy programmes launched by the EU to support recovery, resiliency and new modes of growth are definitively future-oriented. Foresight is therefore particularly appropriate for the current season of EU policies. The European Commission adopted the first-ever Strategic Foresight Report in September 2020 to set out the rationale of foresight and chart priorities for the development of EU policies. In the EU legislative cycle, new policy initiatives are supported by impact assessment. To state that foresight and impact assessment are not foes, but friends is uncontroversial, but less obvious is how to pin down how exactly they can be friends. We explain how foresight can add to impact assessment, and outline some issues that will have to be addressed in the near future.

This handbook is published in the context of AI Watch, the European Commission’s knowledge service to monitor the development, uptake and impact of AI for Europe, which was launched in December 2018 as part of the Coordinated Plan on the Development and Use of Artificial Intelligence Made in Europe. As part of the AI Watch, a specific Task addresses the role of AI for the public sector, and it is set out to provide actionable guidelines for the adoption of AI in the public sector in the EU. The public sector deserves special attention in this regard, as it differs from the private sector in a number of ways and features.

On 21 April 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” (AIA). This article takes a critical look at the proposed regulation. After an introduction (1), the paper analyzes the unclear preemptive effect of the AIA and EU competences (2), the scope of application (3), the prohibited uses of Artificial Intelligence (AI) (4), the provisions on high-risk AI systems (5), the obligations of providers and users (6), the requirements for AI systems with limited risks (7), the enforcement system (8), the relationship of the AIA with the existing legal framework (9), and the regulatory gaps (10). The last section draws some final conclusions (11).

This article will unpack the European Draft Act on Artificial Intelligence (AI), the first (both in Europe and in the world) far-reaching regulation in this domain. It categorizes AI systems in three (eventually four) levels of risk and it assigns to each level a particular legal framework, with its own limitations and obligations. The Draft Act was created for laudable purposes, namely to harmonize digital development with fundamental rights and European values. However, similar to many other ambitious projects, it falls short in many ways. As this article will show, several issues are incompletely regulated, and there are doubts about the exact scope and content of the legal solutions outlined in the draft. It also potentially overlaps with several other European norms, introducing the possibility of conflicts. Finally, the law’s focus on fundamental rights may come at the price of digital innovation, although critics claim that the Draft Act does not go far enough to protect such rights.

Critics raise alarm bells about governmental use of digital algorithms, charging that they are too complex, inscrutable, and prone to bias. A realistic assessment of digital algorithms, though, must acknowledge that government is already driven by algorithms of arguably greater complexity and potential for abuse: the algorithms implicit in human decision-making. The human brain operates algorithmically through complex neural networks. And when humans make collective decisions, they operate via algorithms too—those reflected in legislative, judicial, and administrative processes. Yet these human algorithms undeniably fail and are far from transparent. On an individual level, human decision-making suffers from memory limitations, fatigue, cognitive biases, and racial prejudices, among other problems. On an organizational level, humans succumb to groupthink and free-riding, along with other collective dysfunctionalities. As a result, human decisions will in some cases prove far more problematic than their digital counterparts. Digital algorithms, such as machine learning, can improve governmental performance by facilitating outcomes that are more accurate, timely, and consistent. Still, when deciding whether to deploy digital algorithms to perform tasks currently completed by humans, public officials should proceed with care on a case-by-case basis. They should consider both whether a particular use would satisfy the basic preconditions for successful machine learning and whether it would in fact lead to demonstrable improvements over the status quo. The question about the future of public administration is not whether digital algorithms are perfect. Rather, it is a question about what will work better: human algorithms or digital ones.