Latest / AI Ethics with Fexingo: Bias, Safety, and Responsible Artificial Intelligence
When Your AI Therapist Misses the Warning Signs
Episode 87 of AI Ethics with Fexingo examines the risks of AI-powered mental health chatbots. Lucas and Luna dig into a 2025 study from the Journal of Medical Internet Research that found popular AI therapy apps missed critical suicide indicators in 62% of simulated patient conversations. They explore why these systems fail — from training data that sanitizes crisis language to economic incentives that prioritize engagement over safety — and what it means for the millions who now turn to AI for emotional support. The hosts discuss how companies like Koko and Woebot have handled these…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.