Artificial intelligence (AI) is now essential in almost all fields, including ours, that of mental health. This technological novelty raises a central question: could AI one day replace psychologists? The answer is not so simple to give.
The risks of AI in clinical practice
The Order of Psychologists of Quebec mentions several major risks linked to the use of AI in clinical practice (Risks of AI in clinical practice and mitigation strategies, OPQ, 2024).
The first risk mentioned is anthropomorphism, or the tendency to attribute human qualities to AI such as empathy or clinical judgment (OPQ, 2024). These systems unfortunately rely only on statistical models and not on a real understanding of emotions. AI can also produce hallucinations (false beliefs), i.e. content errors or invented but credible information, which can negatively influence a clinical decision.
The OPQ (2024) also mentions the risk of bias in the data collected, which could reproduce social or cultural prejudices. The ethical issues are also worrying: for example, if an AI makes a mistake, who should assume the consequences? Would it be the responsibility of the designer, the model or the organization?
Why will AI not replace psychologists?
For this reason, the Order recalls that it is essential to keep the psychologist as a support person in need. Artificial intelligence can never be a tool for autonomous decision-making or even processing.
So, to the question “Will AI replace psychologists?” ”, the answer is no (at least, not in the near future). Empathy, clinical creativity and human understanding cannot be replaced by a robot. Artificial intelligence will be able to automate some repetitive tasks, but the therapeutic relationship remains irreplaceable.
##Source
Order of Psychologists of Quebec. (2024). Risks of AI in clinical practice and mitigation strategies. https://www.ordrepsy.qc.ca/-/risks-de-l-ia-dans-la-pratique-clinique-et-strategies-d-attenuation