Think your AI chatbot is helping you? Researchers say there’s a risk users may not recognize

Despite repeated warnings about the risks of artificial intelligence, users have become increasingly reliant on AI chatbots—particularly those offering companionship. Yet despite chatbots’ utility for tasks like answering quick queries or offering brief advice, their benefits might not be outweighing the costs.

A new, peer-reviewed study from researchers at Northeastern University’s Network Science Institute suggests that AI chatbots used for companionship, despite offering momentary benefits, are causing their users long-term damage and psychological harm.

The study dives into a common phenomenon where users turn to AI companion platforms for services like emotional support and well-being. 

The study looked at both purpose-specific AI chatbots, like Replika and Character.AI, and the popular chatbots like OpenAI’s ChatGPT or Anthropic’s Claude. 

Across the various models, the story is often similar: A user turns to AI for advice and AI returns a sycophantic reply engaging the user even more, particularly as the person begins to anthropomorphize the technology.

“The fluent and apparently empathic responses from these models lead users to engage with them not only as tools but also as if they were social entities,” the study says. “This framing is conceptually misleading and may pose risks across different user profiles, particularly for vulnerable individuals.”

The study, which looked at reports for both clinical and nonclinical populations, found some evidence of short-term benefits like a brief reduction in loneliness and improvement in one’s mood. Yet longer term, the effects turned to emotional dependence, delusional ideation, and sometimes even self-harm.

The study is not the first evidence of the possible adverse effects of relying on AI chatbots. The consequences of users turning to chatbots for emotional support rather than seeking out experienced professionals have gained particular attention in recent years, following a series of high-profile cases in which teenagers died by suicide and their parents blamed AI chatbots.