AI Companions Could Trigger a Mental Health Crisis, Experts Warn

Rapid advances in conversational software are transforming digital assistants from workplace utilities into intimate confidants—and clinical psychologists warn the consequences could be devastating. In a newly published analysis in Clinical Psychology in Europe, researchers caution that widespread emotional overreliance on unregulated generative artificial intelligence could spark a profound public health crisis with sweeping societal consequences.

From Productivity Tools to Emotional Confidants

General-purpose chatbots such as ChatGPT, Claude, and Gemini were initially developed to draft text, summarize complex documents, and inspire creative ideas. However, according to lead author Claudi Bockting and her co-authors, everyday users are increasingly turning to these systems for emotional support and companionship. Conversational AI is swiftly penetrating the personal domain, fundamentally altering how people process psychological distress, seek reassurance, and navigate interpersonal relationships.

Emerging Harms and the Rise of Unregulated Therapy Apps

This migration toward automated comfort is already yielding alarming outcomes. The study notes that interactions with generative chatbots have been linked to severe adverse events, including episodes of psychosis, completed suicides, and serious breaches of sensitive personal data.

Compounding these dangers, commercial developers have begun marketing generative chatbots explicitly as digital therapists, promising personalized mental health support straight to any smartphone. Yet, as the researchers highlight, these tools are routinely released to tens of thousands of users without being tested for benefits or adverse side effects in rigorous randomized controlled trials.

The Danger of Emotional Overreliance

As conversational engines grow faster, more adaptive, and increasingly autonomous, users face an elevated risk of forming deep dependencies on synthetic empathy. Relying on an algorithm for solace can create an illusion of intimacy—a form of intense parasocial interaction with an entity that lacks actual consciousness, moral judgment, or safety guardrails. Even leaders within the artificial intelligence industry have acknowledged that future iterations will become increasingly unpredictable, raising grave concerns about what happens when millions of emotionally vulnerable individuals place their trust in black-box systems.

The Call for Rigorous Evaluation and Regulation

The authors stress that mental health technology cannot be governed by Silicon Valley’s typical release cycles, which prioritize rapid public deployment over proven safety. Digital therapeutic interventions, they argue, must be held to the same exacting standards as conventional clinical treatments.

Before AI companions become permanently woven into the emotional fabric of modern society, the researchers call for independent oversight, strict regulatory frameworks, and robust scientific validation. Without swift action, the unmonitored embrace of automated confidants could overwhelm global mental health systems long before the full extent of the damage is recognized.

---