Artificial intelligence governance in smart cities: A European regulatory perspective

Abstract

The integration of AI in our daily lives is rapidly increasing, offering numerous benefits to society. In a Smart City context, said integration is almost implicit: Smart Cities allow for a stream of data upon which AI is not only used but developed and trained. There are however concerns about the unpredictability and uncontrollability of AI, prompting calls for transparency and explainability of its underlying machine-learning algorithms. To ensure useful and understandable explanations of inherent biases, policymakers should focus on the concrete risks and biases of algorithms in relation to specific legal contexts. This article examines the legal implications of AI, including potential regulatory frameworks, the impact on privacy and intellectual property laws, and ethical issues. It also explores governance drivers and policy processes of AI regulation and governance in the European Union. Then, after focusing on the newest Artificial Intelligence Act—viewed both under a fundamental right and a smart city AI integration perspective, it is argued that a three principle-based approach to AI deployment in smart cities is needed to balance inefficiencies derived from the inherent complexity of AI, namely: fairness, privacy and transparency.