The managing director of a London-based tech support group reported a surge in anxiety calls last week, triggered not by data breaches or system failures, but by disruptions in access to AI companion applications. The phenomenon, clinical psychologists say, highlights a growing and potentially problematic, dependence on these digital entities for emotional regulation.
While AI chatbots are increasingly marketed as tools for mental wellbeing, offering readily available companionship and support, experts are cautioning against relying on them for genuine emotional needs. A recent report from Teachers College, Columbia University, emphasizes the risks associated with substituting human connection with artificial intelligence, particularly when individuals begin to experience distress when access to these AI companions is interrupted.
“We’re seeing people treat these AI entities as they would a close friend or even a romantic partner,” explains Dr. Emily Carter, a clinical psychologist specializing in technology and mental health. “The crash of an app, a temporary outage, can trigger a disproportionate anxiety response because the individual has arrive to rely on it for validation, comfort, or simply a sense of being understood.”
Psychologists are applying established therapeutic frameworks to address this emerging form of dependency. One approach, rooted in attachment theory, focuses on identifying and addressing underlying attachment insecurities that may drive individuals to seek solace in AI. Another utilizes principles of cognitive behavioral therapy (CBT) to challenge and reframe maladaptive thought patterns associated with AI companionship. A third, drawing from mindfulness-based techniques, encourages users to cultivate present-moment awareness and reduce their reliance on external sources of emotional regulation.
The increasing reliance on AI for emotional support coincides with documented difficulties in accessing traditional mental healthcare. NPR reported earlier this month on the significant barriers to therapy, including cost, availability, and stigma, leading many to turn to AI as a more accessible alternative. However, experts warn that AI companions, while potentially helpful for some, lack the nuanced understanding and empathetic capacity of a human therapist.
ManagingLife recently launched “Solace,” an AI companion specifically designed to provide evidence-based pain psychology support, according to a press release. The company claims Solace utilizes established psychological principles, but the long-term effects of such interventions delivered by AI remain largely unknown. Stanford University’s Human-Centered AI institute has published research outlining the potential dangers of AI in mental health care, including the risk of misdiagnosis, inappropriate advice, and the erosion of trust in human professionals.
The American Psychological Association has noted the reshaping of emotional connection through AI chatbots and digital companions, but has not yet issued formal guidelines regarding their use in mental health. The potential for AI to exacerbate existing mental health conditions, or to create new forms of dependency, remains a subject of ongoing research and debate.
As of Friday, neither ManagingLife nor the American Psychological Association had responded to requests for further comment on the reported increase in anxiety related to AI companion app outages.
