AI needs a clinical safety check

As the number of mental health AI chatbots grows, Numerof and Associates' Michael Abrams warns that without rigorous testing and validation these largely unregulated tools may pose real risks to patients.