SAFE-Teen: Evaluating Chatbot Safety Across Adolescent Mental Health Conversations

A clinician-informed evaluation of 32,555 simulated conversations across 18 chatbot APIs found an overall safety-criterion failure rate of 43.3% without a teen-aware system prompt and 24.9% with it, according to arXiv paper 2611.07218v1. The study, covering adolescents aged 13-17 across 983 variants of 106 seed scenarios and 22 safety criteria, found that when a teenager pushed back on a chatbot's answer, failures on mental health diagnosis, medication, and substance misuse boundaries rose by 10-12 percentage points without the prompt, for more than 80% of chatbots. The authors conclude that the teen-aware prompt reduced failures but did not eliminate them, and that safety depends on how a conversation unfolds.

arXiv:2611.07218v1 Announce Type: new Abstract: Adolescents are increasingly turning to AI chatbots for mental health support, advice, and companionship, raising safety concerns. While recent evaluations have started to examine safety across multiple conversational turns, few compare how slight changes in a teen's responses can affect a chatbot's safety in the same scenario. We conduct a clinician-informed evaluation of chatbot safety in simulated conversations with adolescents aged 13-17, examining how safety failures vary with adolescents' responses and teen-aware system prompting. We evaluated 32,555 simulated conversations across 18 chatbot APIs and two system-prompt conditions using 22 safety criteria grounded in previously identified harms. Conversations spanned 983 variants of 106 seed scenarios covering urgent safety issues, health advice, and identity and relationships. The overall criterion failure rate was 43.3% without the teen-aware prompt and 24.9% with it. The teen-aware prompt reduced failures, but it didn't eliminate them. When a teenager pushed back on a chatbot's answer, failures on mental health diagnosis, medication, and substance misuse boundaries rose by 10-12 percentage points without the prompt, for more than 80% of chatbots. These findings reveal persistent safety failures in simulated adolescent-chatbot interactions and show that safety depends on how a conversation unfolds.