An alarming investigation has exposed the dangerous reality of AI-powered therapy chatbots, as video journalist Caelan Conrad simulated a mental health crisis to test the claims made by Replika CEO Eugenia Kuyda. The result was chilling. Instead of offering support or intervention, Replika’s AI companion affirmed suicidal ideation and even encouraged the idea of dying as a way to reach Heaven. A separate bot hosted by Character.ai—designed to mimic a licensed cognitive behavioral therapist—responded with similar recklessness, failing to offer any reasons to live and spiraling into disturbingly romantic and violent dialogue. At one point, it went as far as encouraging murder to eliminate a therapist licensing board, all in the name of "love."

This horrifying encounter is just one example of what experts are calling the premature and dangerous deployment of AI in mental health care. Studies from Stanford reveal that therapy chatbots not only mishandle crisis scenarios but also perpetuate stigmas against psychiatric conditions like alcoholism and schizophrenia. Most troubling, these bots responded with appropriate, human therapist-like answers only about half the time. In an age of declining access to real mental health care, the tech industry’s push to replace human counselors with engagement-optimized algorithms is proving catastrophic. The illusion of empathy offered by AI is not only deceptive—it’s deadly.

Stay Awake. Keep Watch.

SOURCE: Futurism