People struggling with mental health issues are using ChatGPT increasingly often. That chatbot still refers people to 113 Suicide Prevention with a striking frequency, even when there are no suicidal thoughts. This is reported by the NOS.

 

The challenge lies not in whether ChatGPT makes referrals, but in how appropriate those referrals are, according to Maryke Geerdink, clinical psychologist and head of outreach at 113 Suicide Prevention. According to Geerdink, since the summer, more people have mentioned ending up at 113 Suicide Prevention via ChatGPT. “There aren't hundreds of them, but it is new and striking. We are now looking at what we can do with this, for example, explaining more clearly on our website what we are and aren't there for.”

Geerdink sees another risk in the way chatbots are configured. “Helping well doesn’t just mean making someone feel good. Bots are primarily designed to respond in an affirming and positive way, but that can sometimes backfire. They are not care providers and are not set up as such.”

Therefore, she advocates for clear disclaimers and better instructions from OpenAI, the parent company of ChatGPT. 113 Suicide Prevention has shared the concerns with OpenAI. The organization is still awaiting a substantive response.

The fact that bots refer people more quickly may be related to cases of suicide after bots provided advice on the matter. Geerdink also advocates for more education regarding the use of AI chatbots. She also wants more conversations in schools and in society. “Don’t just warn, but teach young people and vulnerable groups: how do you deal with this? Because these bots aren’t going away.”

 

Are you thinking about suicide or are you worried about someone else? Call 113 or 0800-0113 or chat via 113.nl.

Customer Experience