Table 1 AI involvement in psychiatry: a selection of ethical challenges

Prevention

Harm prevention/safety [21], efficacy [22], data security [23]

Suicide prevention (e.g. based on the electronic health records) [24]

Diagnosis prognosis

Transparence (“black box”), data privacy [4], patient autonomy [18, 19]

Use of complex/multidimensional biomarkers [22], use of diagnostic apps

Therapy

Lack of human contact, safety [19]

Chatbots/robots as “therapists” [24]

Education

Responsibility [23], lack of guidance and training [21]

Digital competence in the field of AI necessary

Research

Participants’ data security, informed consent, data privacy, safety [25]

Study participation, data collection (e.g. monitoring)