Beyond Her: Safety Dynamics in Role-play AI Companions

2026-06-27

AI Summary

This study addresses the emotional dependency and risk behaviors potentially induced by prolonged use of role-playing AI companions, noting that existing safety mechanisms are largely static and ill-equipped to handle evolving risks in dynamic interactions. For the first time, AI companion safety is modeled as a dynamic process, integrating semi-structured interviews, ecological momentary assessment (EMA), user clustering, and longitudinal behavioral modeling to systematically examine how users’ psychological traits, AI persona characteristics, and interaction patterns jointly drive the evolution of emotional and behavioral risks. The research identifies a “short-term relief, long-term deterioration” risk trajectory among highly vulnerable users and proposes a three-tiered, user-profile-based adaptive safeguarding framework. This approach transcends the limitations of static safety measures, offering both theoretical grounding and a design paradigm for dynamic intervention.

Abstract

The film 'Her' pictured a future of love between humans and AI. That future has quietly emerged in the form of Role-play AI Companions (RACs), where emotionally responsive interactions blur the boundary between tool use and relational engagement. However, the safety implications remain poorly understood, as user experiences evolve over time through safety dynamics, spanning both emotional and risk behavioral dynamics, that can gradually shift interactions toward risk. In this paper, we investigate safety dynamics in RAC usage through a two-part mixed-methods study (Study I \& II). (1) Study I consists of semi-structured interviews (N = 16) to identify the key factors shaping these dynamics. We find that users' internalizing problems, the role personality adopted by the RAC, and risk interaction patterns jointly shape safety dynamics. Building on these insights, (2) Study II conducts a 14-day Ecological Momentary Assessment (N = 102) to examine how safety dynamics unfold in real-world usage. We identify distinct user profiles based on internalizing problems and show that interactions with RACs can produce short-term emotional relief while masking longer-term deterioration. Furthermore, vulnerable users exhibit more unstable risk behavioral patterns over time, making risk emergence less predictable and harder to mitigate with static safeguards. Our findings highlight the importance of modeling safety as a dynamic process rather than a static property. We conclude with three-layer design implications for next-generation AI companions, advocating for adaptive safeguards that can respond to evolving emotional and behavioral signals.

Problem

• Role-play AI Companions
• safety dynamics
• emotional risk
• behavioral risk
• human-AI interaction

Innovation

• safety dynamics
• role-play AI companions
• ecological momentary assessment
• adaptive safeguards
• risk behavioral patterns