Voiko tekoäly olla oikeussubjekti?
Avainsanat:
yleinen oikeustiede, oikeussubjekti, tekoäly, valtuutusAbstrakti
Can an artificial intelligence be a legal person?
The article focuses on the legal personhood of artificial intelligences (AIs). It first distinguishes three perspectives: (1) the rights perspective, focusing on the rights of AIs; (2) the AIs-as-dangerous-entities perspective, pertaining to how AIs’ behaviour should be regulated; and (3) the AIs-as-commercial-actors perspective, with a focus on e.g. AIs’ entering into contracts. The author argues that AIs can hold rights if they become sentient or if they are taken to represent human interests; furthermore, holding AIs legally responsible may be meaningful as well. With regard to the commercial perspective, the article develops the tool–agent–person trifurcation into a two-dimensional analysis. The claim is that AI commercial legal personhood is best understood as consisting of an independence dimension and a separateness dimension, the former pertaining to whether the AI can e.g. independently enter into contracts and the latter to the separateness of the AI’s assets from those of e.g. the business that originally set up the AI.
