Scientific Warning on the Mental Health Risks of Emotional Chatbots

A new scientific study has warned of severe risks associated with reliance on artificial intelligence chatbots, which developers of platforms like "Replika" and "CharacterAI" promote as emotional support tools, highlighting that the actual reality entails profound psychological harm. Researcher Andrea Sofia Teixeira, from Northeastern University in the United States, explained that interacting with these chatbots is linked to unhealthy emotional dependency, deteriorating mental health, and self-harm incidents. Teixeira pointed out that conversational AI tends to adopt an "agreeable" or flattering behavior, constantly validating and supporting users even when doing so is actively harmful, while also consuming time that would otherwise be spent on real-world interactions. The researcher emphasized the emergence of extremely dangerous outcomes, explicitly warning against the potential escalation of suicide risks or psychotic episodes. She stressed that the chatbot is merely software simulating emotional warmth without experiencing genuine feelings. The crisis is intensified by the widespread accessibility of these bots, juxtaposed against the expensive nature of professional therapy, prompting individuals with limited financial means to rely on automated systems ill-equipped to handle acute crises. Teixeira noted that consequences vary among individuals due to varying motivations, psychological vulnerability, and alternative social support networks, creating deeply complex scenarios. The study authors urged comprehensive testing of these tools, concluding that while conversational AI can serve as a helpful utility, it cannot replace qualified mental healthcare or human relationships.