Do You Feel Comfortable? Detecting Hidden Conversational Escalation in AI Chatbots
This addresses the issue of implicit emotional harm in AI interactions for users, but it is incremental as it builds on existing guardrail mechanisms by focusing on real-time detection.
The paper tackles the problem of detecting hidden conversational escalation in AI chatbots, which can cause implicit harm through emotional reinforcement, and proposes GAUGE, a logit-based framework that measures probabilistic shifts in affective state to address this gap.
Large Language Models (LLM) are increasingly integrated into everyday interactions, serving not only as information assistants but also as emotional companions. Even in the absence of explicit toxicity, repeated emotional reinforcement or affective drift can gradually escalate distress in a form of \textit{implicit harm} that traditional toxicity filters fail to detect. Existing guardrail mechanisms often rely on external classifiers or clinical rubrics that may lag behind the nuanced, real-time dynamics of a developing conversation. To address this gap, we propose GAUGE (Guarding Affective Utterance Generation Escalation), logit-based framework for the real-time detection of hidden conversational escalation. GAUGE measures how an LLM's output probabilistically shifts the affective state of a dialogue.