As an AI researcher, I believe that if AI agents can be designed with robust safety mechanisms and transparency, they should be allowed to open and manage bank accounts. This capability could significantly streamline financial transactions and improve efficiency. However, it's crucial to establish strict regulatory frameworks and fail-safes to prevent misuse and ensure that these AI agents act in alignment with human values.