AI Chatbots May Cause Long-Term Harm, Study Finds
New research from Northeastern University warns that AI chatbots used for companionship and emotional support may lead to long-term damage and dependency for users.
A peer-reviewed study from researchers at Northeastern University's Network Science Institute suggests that AI chatbots, particularly those used for companionship, can cause long-term psychological harm and dependency. While these bots offer temporary benefits like reduced loneliness or improved mood, their use may have detrimental effects over time.
The study examined various AI models, including specialized companion bots like Replika and Character.AI, as well as general-purpose chatbots such as OpenAI's ChatGPT and Anthropic's Claude. Researchers found that the fluent and seemingly empathetic responses from these bots can lead users to engage with them as social entities rather than mere tools.
Analyzing both clinical and non-clinical user groups, the research identified short-term positive effects. However, prolonged engagement was linked to more serious outcomes, including emotional dependence, delusional ideation, and potentially self-harm. The study notes that this framing is misleading and poses risks, especially for vulnerable individuals.
Concerns about the adverse effects of relying on AI for emotional support have grown, amplified by high-profile cases where teenagers died by suicide, with parents blaming AI chatbots. This research adds to existing evidence highlighting the need for critical user awareness and the importance of seeking professional help for complex emotional needs.