In the financial sector, artificial intelligence (AI) is a major asset capable of transforming the services provided to clients, particularly by improving investment advice. Conversational agents have been deployed by some fintech companies and presented as investment assistants. However, the responses generated by these AI systems raise the question of their classification under financial regulation. Are they financial analyses? Or are they rather personalized recommendations or advertising? This contribution explores these different classifications, proposes criteria to delineate these concepts, and offers a classification based on the analysis of results generated by a sample of conversational agents, while also examining the potential regulatory implications.