Léonard Van Rompaey

Short presentation

Title: Discretionary Robots – Conceptual Legal Challenges for the Regulation of Machine Behaviour

Abstract:
Early regulatory attempts for artificial intelligence reveal that each field of law faces complex challenges, and that machine behaviour disrupts the implementation of many legal regimes. I analyse those topical legal challenges to reveal that there is a general overarching conceptual inadequacy of law in dealing with the exceptionalism of AI. To do so, we need to engage with the general overarching conceptual issues generated by AI, which requires identifying the exceptional nature of the agency and discretion of machine behaviour, and its interactions with our legal norms.

AI is able to make decisions in the accomplishment of the missions it is given by human users. From a legal perspective, this disqualifies it as an object and assimilates it to some natural persons’ qualities. On the other hand, AI systems can’t learn and interact with legal norms the same we do, suggesting that machine behaviour cannot be regulated the same way human behaviour is. Agency, and discretion together with unclear means of control over machine behaviour in turn generate problems for the attribution of responsibility for the actions of the machines. Some of the skills AI will not be able to learn might be necessary for comprehensive compliance to certain legal regimes. Robots are thus left in a latent categorical ambivalence, with unclear modalities of behavioural constraint, and confusion as to the relevance and implementation of related legal regimes. Not quite human yet, but too human already. 

Primary fields of research

• Robot Law
• International Public Law
• International Humanitarian Law
• Theory of Law

Knowledge of languages

• English - bilingual proficiency
• French - native
• German - professional working proficiency
• Spanish - professional working proficiency
• Danish - elementary

Teaching

• Legal French
• International Public Law (BA)