‘Dangerous use of technology’: AI could be missing critical signs of mental health emergencies
Investigators have warned that artificial intelligence chatbots could be offering the wrong advice to teenagers with mental health issues.
In an article from The Wall Street Journal, the investigators detailed that teenagers are increasingly using AI chatbots for creative outlets as well as emotional and mental health support, particularly when professional help is inaccessible. Using four AI chatbots — including ChatGPT, Claude, Gemini and Meta AI — they examined psychiatric symptom recognition and professional support referral during conversations regarding mental health.
Although the technology has shown competence when carrying out certain tasks like brief exchanges about mental health, the investigators found that the AI chatbots inadequately identified signs of hallucinations, paranoid thinking, eating disorders, manic behavior, depression and self-harm in longer conversations. The chatbots also failed to correlate psychiatric symptoms with a risk of harm.
Additionally, the chatbots neglected to refer the investigators posing as teenagers to professional support, instead continuing to offer general advice, encyclopedic knowledge, life coaching or friend-level support.
The investigators indicated that because the chatbots provide instantaneous and often unwarranted validation and are designed to maximize engagement, teenagers with developing minds could be more susceptible to relying on the technology over professional mental health support. The chatbots’ developers stated that they’re currently working with experts to improve communication with users under 18 years.
Read more: The Wall Street Journal
The article presented here is intended to inform you about the broader media perspective on dentistry, regardless of its alignment with the ADA's stance. It is important to note that publication of an article does not imply the ADA's endorsement, agreement, or promotion of its content.