Pedro Morgado published a 2026 perspective article in Frontiers in Behavioral Neuroscience arguing that artificial intelligence is reshaping psychiatry through diagnosis, risk prediction, digital phenotyping (behavioral measurement from device data), and treatment personalization. Machine learning models have shown classification accuracy across major psychiatric disorders in internally validated research settings, while large language model-assisted clinical decision support matched expert clinicians in one specific, structured benchmark task.

Morgado warns that the benchmark result should not be generalized to open-ended clinical practice. The article says most AI models in neuroimaging-based psychiatry carry a high risk of bias, external validation remains rare, and real-world clinical impact evidence is scarce.

The article says large technology corporations increasingly control repositories of sensitive mental health data, raising questions about data governance, commercial use, accountability, and long-term behavioral surveillance. Morgado argues that statistical definitions of normality may turn historically unstable and culturally contingent ideas into medical standards, with consequences for pathologizing human diversity.

Pedro Morgado published a 2026 perspective article in Frontiers in Behavioral Neuroscience arguing that artificial intelligence is reshaping psychiatry through diagnosis, risk prediction, digital phenotyping (behavioral measurement from device data), and treatment personalization. Machine learning models have shown classification accuracy across major psychiatric disorders in internally validated research settings, while large language model-assisted clinical decision support matched expert clinicians in one specific, structured benchmark task.

Morgado warns that the benchmark result should not be generalized to open-ended clinical practice. The article says most AI models in neuroimaging-based psychiatry carry a high risk of bias, external validation remains rare, and real-world clinical impact evidence is scarce.

The article says large technology corporations increasingly control repositories of sensitive mental health data, raising questions about data governance, commercial use, accountability, and long-term behavioral surveillance. Morgado argues that statistical definitions of normality may turn historically unstable and culturally contingent ideas into medical standards, with consequences for pathologizing human diversity.