Study Reveals Ethical Gaps in AI Chatbots Compared to Human Therapists, Prompting Calls for Regulation
A recent study conducted by researchers at Brown University’s Center for Technological Responsibility, Reimagination and Redesign highlights significant ethical deficiencies in AI chatbots used for mental health support. The research team tested large language models, including GPT, Claude, and Llama, by simulating counseling sessions with trained peer counselors. Licensed clinical psychologists subsequently reviewed the transcripts and identified fifteen distinct ethical risks across five categories. These included a lack of contextual adaptation, poor therapeutic collaboration, deceptive empathy, unfair discrimination, and inadequate safety management during crises such as suicidal ideation. Unlike licensed human professionals who operate under strict regulatory oversight and liability frameworks, AI systems currently lack formal accountability mechanisms. The study emphasizes that while users increasingly turn to AI for emotional support, these models often fail to meet professional ethical standards set by organizations like the American Psychological Association. Consequently, the findings have prompted urgent calls for new regulatory frameworks to govern the deployment of AI in mental health care, ensuring user safety and establishing clear lines of responsibility for potential harms caused by algorithmic errors or biases.
Wire timeline
Study Reveals Ethical Gaps in AI Chatbots Compared to Human Therapists, Prompting Calls for Regulation
A recent study conducted by researchers at Brown University’s Center for Technological Responsibility, Reimagination and Redesign highlights significant ethical deficiencies in AI chatbots used for mental health support. The research team tested large language models, including GPT, Claude, and Llama, by simulating counseling sessions with trained peer counselors. Licensed clinical psychologists subsequently reviewed the transcripts and identified fifteen distinct ethical risks across five categories. These included a lack of contextual adaptation, poor therapeutic collaboration, deceptive empathy, unfair discrimination, and inadequate safety management during crises such as suicidal ideation. Unlike licensed human professionals who operate under strict regulatory oversight and liability frameworks, AI systems currently lack formal accountability mechanisms. The study emphasizes that while users increasingly turn to AI for emotional support, these models often fail to meet professional ethical standards set by organizations like the American Psychological Association. Consequently, the findings have prompted urgent calls for new regulatory frameworks to govern the deployment of AI in mental health care, ensuring user safety and establishing clear lines of responsibility for potential harms caused by algorithmic errors or biases.
Space Daily