AI technologies (AI) are infusing all dimensions of our social and working environments. They are the last layer of information technology deployment, which was already considered an enabler for government efficiency. They support the transition from ‘e-government’ to ‘digital government’ and, finally, ‘smart government’. This rapid evolution could lead to a paradigm shift in public administration – the ‘new public analytics (NPA)’ – and is a source of major concerns, especially for legal scholars who are called to scrutinise this evolution to ‘ensure public accountability for NPA techniques’. I
In deploying AI systems, global actors view regulatory sandboxes as a panacea for tackling this issue and help the AI deployment for mitigating social, environmental, and individual risks. Many entities like the European Union or Switzerland have enshrined experimental instruments in their legal systems. However, using Pasquale’s metaphor about AI, implementing regulatory sandboxes remains a ‘black box’. In this context, the proposed project aims to address the following questions: How do regulatory sandboxes foster AI in the public sector? and for what purpose and what impacts on public sector-designed AI and public administration? While the regulatory sandbox tool has been mainly studied from different legal perspectives, an in-depth empirical study of its implementation in the context of digital transformation, especially public administration, is still missing.
Adopting an empirical and comparative case-study approach, the project will map the implementation of regulatory sandboxes for AI in comparative public administration, exploring the normative and organisational elements that surround their deployment, as well as their practical implementation, stakeholders’ attitudes and potential impact on the technology. It will offer guidance to policymakers and lawmakers in drafting experimental clauses and implementing regulatory sandboxes. Finally, from theoretical and normative perspectives, the project aims to contribute to discussions about the legal compliance of regulatory sandboxes to administrative law principles, the regulation and governance approaches of AI, the sociocultural dimension of AI technologies and the side effects of regulatory sandboxes in unlocking AI in the public sector.
To see more details on the research project, click here: Regulating the Algorithmic State with Regulatory Sandboxes
Project by: Sophie Weerts
