Description
The development of artificial intelligence (AI) has challenged legal instruments and scholars everywhere: how to fit this new reality in existing frameworks and contexts? AI systems can make decisions with legal consequences - just like a human would - but is in itself an object that is not receptive to the same incentives and emotions - unlike humans. In its core, AI systems are very akin to software, yet have more capabilities than many other, more traditional, software. AI systems can be regulated (e.g. European Union AI Act, Regulation 2024/1689) but it has no rights or duties. Whereas a falling chimney, a car that does not brake and a chemical substance catching fire spontaneously are attributed to actors such as legal persons and humans, the actions and omissions of an AI system can hardly be attributed to these actors because of a lack of control over these actions and omissions.It has been tried often: fitting AI systems and their ‘conduct’ into existing legal frameworks, seeking solutions such as legal personhood. This leads to friction between the old and the new: legal personhood often serves as a way of protection (e.g. assigning legal personhood to rivers so they can be represented in court to claim protection from polluters) whereas legal personhood of AI systems might also lead to duties of the AI system. For instance the duty to pay compensation for damage caused by the actions and omissions of the AI system. What rights would an AI system have and what would this mean in practice - would an AI system have a bank account for its wage, so it can pay damages from its savings? - is up for debate. This contribution aims to explore an alternative approach, by introducing the notion of artificial conduct as legally relevant conduct, that could become subject to regulation. This notion can fill the gap that currently exists in many legal systems that focus on human conduct and conduct of a legal person in the conventional sense (i.e. a company). Artificial conduct would encompass the actions and omissions of AI systems, the conduct that is hard to control by any traditional actor. In this contribution, a definition for artificial conduct is developed and the potential of artificial conduct as a subject for regulation is explored. Can the notion of artificial conduct overcome the current challenges in regulating AI systems and their use?| Period | 16-Apr-2026 |
|---|---|
| Event title | Bileta 2026: “Technology is a useful servant but a dangerous master” |
| Event type | Conference |
| Location | Aberystwyth , United KingdomShow on map |
| Degree of Recognition | International |