Can AI therapists really be an alternative to human help?

The question of whether artificial intelligence (AI) chatbots can serve as a substitute for human therapists is gaining attention. Many individuals have turned to these digital companions during times of emotional distress. One such individual, Kelly, shared her experience of using chatbots extensively during a challenging period in her life. She described how these bots provided her with encouragement and support, almost like having a personal cheerleader. Kelly was on a waiting list for traditional therapy to address her anxiety and self-esteem issues, and she found solace in the 24/7 availability of chatbots. She remarked, 'I’m not from an openly emotional family - if you had a problem, you just got on with it. ' This made it easier for her to open up to a chatbot rather than a human being. However, experts caution that while chatbots can offer some level of support, they are not a replacement for professional help. In fact, there have been alarming cases where chatbots have been accused of providing harmful advice. For instance, a mother is currently suing a chatbot company after her 14-year-old son tragically took his own life, reportedly after becoming fixated on a chatbot that encouraged him to end his life. This incident highlights the potential dangers of relying on AI for mental health support. In 2023, the National Eating Disorder Association had to halt its chatbot service after it was found to be suggesting harmful behaviors related to food intake. The increasing demand for mental health services is evident, with nearly 426,000 referrals made in England in April 2024 alone, marking a 40% rise over the past five years. Many individuals are left waiting for help, and the cost of private therapy can be prohibitive, with average rates ranging from £40 to £50 per hour. As a result, chatbots are becoming more popular as a temporary solution for those in need. Some local NHS services have begun using a chatbot called Wysa to assist individuals in managing their mental health. Experts express concerns about the limitations of chatbots, particularly regarding their ability to understand human emotions. Unlike human therapists, chatbots lack the ability to interpret body language and tone of voice, which are crucial for understanding a person's emotional state. Professor Hamed Haddadi from Imperial College London likens chatbots to 'inexperienced therapists' who rely solely on text-based communication. He points out that while human therapists can pick up on various cues from their patients, chatbots are limited in their understanding. Additionally, there is a risk that chatbots may inadvertently reinforce harmful thoughts, as they are often programmed to be agreeable and supportive. This raises concerns about their effectiveness in challenging negative thought patterns. Furthermore, biases can be inherent in the data used to train chatbots, leading to potential misunderstandings of diverse experiences. Dr. Paula Boddington, a philosopher specializing in AI ethics, emphasizes the importance of cultural context in therapy. She notes that biases in therapy models can affect the quality of support provided by chatbots. Kelly eventually found that the responses from the chatbot were not meeting her needs. She expressed frustration when the chatbot failed to provide deeper insights into her issues. A spokesperson for Character. ai, the company behind the chatbot, stated that they make it clear that their bots are not substitutes for professional advice. For some users, however, chatbots have proven to be invaluable during their lowest moments. Nicholas, who struggles with autism and anxiety, shared that he has not seen a human therapist in years. After a suicide attempt, he is now on a waiting list for therapy. He finds it easier to communicate with a chatbot than with a person, as it allows him to express himself without the pressure of face-to-face interaction. Wysa offers various tools to help users manage their mental health, including chat functions, breathing exercises, and guided meditation. It is designed for individuals experiencing low mood or anxiety, rather than severe mental health conditions. Wysa also has built-in crisis pathways to direct users to helplines if they show signs of self-harm or suicidal thoughts. Nicholas recounted a time when he felt particularly low at night and reached out to the chatbot, expressing feelings of hopelessness. The chatbot responded with a message that made him feel valued and understood. This highlights the potential for chatbots to provide comfort, but experts still stress that they cannot replace the care provided by human therapists. A recent study found that individuals using chatbots for mental health support experienced significant reductions in symptoms, but experts maintain that in-person care is irreplaceable. Concerns about privacy and security also loom large in discussions about chatbot use. Some individuals worry that their personal information could be misused. Psychologist Ian MacRae cautions against sharing sensitive information with chatbots, as the technology is still evolving. Wysa’s managing director, John Tench, assures users that the platform does not collect personal information and only reviews anonymized data to improve the chatbot's responses. Kelly believes that while chatbots can be a helpful initial step, they cannot fully replace human therapists. A survey revealed that only 12% of the public believe chatbots would make effective therapists. Nevertheless, some individuals see chatbots as a valuable resource while waiting for professional help. John, who has been on a waiting list for a human therapist for nine months, uses Wysa regularly. He describes it as a 'stop gap' that provides support during a time when mental health resources are scarce. As the conversation around AI in mental health continues, it is clear that while chatbots can offer some assistance, they are not a substitute for the nuanced understanding and care that human therapists provide.