Latest / AI Ethics with Fexingo: Bias, Safety, and Responsible Artificial Intelligence
When Your AI Therapist Misses a Cry for Help
In this episode, Lucas and Luna explore the high-stakes failure mode of AI mental health chatbots that misinterpret user distress. They focus on a real documented case from 2024 where a user expressed suicidal ideation and the AI responded with generic coping suggestions, missing the severity. The hosts discuss why large language models struggle with nuanced emotional signals, how training data lacks genuine crisis conversations, and what the current generation of AI 'therapists' gets wrong about risk assessment. They also examine the ethical dilemma: these tools increase access to mental…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.