The debate on the legal status of artificial intelligence and robots incor-porating it has grown significantly in recent years. Although isolated cases and previous reflections by the European Parliament have hypothesized forms of “electronic personality”, the current orientation of the European Union treats even the most advanced AI systems as mere products, albeit at different levels of risk. However, the question of the legal subjectivity of robots remains open and continues to stimulate broader reflections. This contribution situates this debate in the historical context of the evolution of legal personality, showing how scientific developments, particularly in biomedicine, have already led to the recognition of limited legal subjectivity before birth and after death. Similarly, the digital and AI revolutions now offer solid reasons to rethink traditional legal categories. After reconstructing the current regulatory framework, which continues to treat AI as a manufactured product, the article critically explores the possibility of recognizing robots as having a distinct legal status, assessing the implications and potential conflicts with the subjectivity already recognized for natural persons. In conclusion, the article proposes introducing a new legal category, “legal agency”, suitable for formally recognizing a special legal status for the most advanced AI systems without equating them with human beings.