Abstract: The diagnosis and treatment of depression currently face significant challenges, with a serious imbalance between the demand and supply of mental health services. Artificial intelligence-based conversational agents offer a promising solution to this issue. However, the development of this technology also brings certain limitations, raising concerns about the potential negative impacts on users’ health. This paper explores the five key ethical challenges faced by conversational agents in the diagnosis and treatment of depression: accuracy, data security and privacy, inherent biases, emotional divide, and patient dependency. It also proposes governance strategies from the perspectives of review and evaluation, data ethics, role positioning, human-machine interaction, and regulation and trust, aiming to ensure that conversational agents improve patient health while effectively mitigating potential risks and promoting their safe and ethical application. 

 Key Words: Conversational agent; Artificial intelligence; Mental health; Depression; Ethics

See the full text at: https://jdn.ucas.ac.cn/public/uploads/files/68832f56b943d.pdf