This year, shocking reports of the disastrous performance of AI chatbots, particularly ChatGPT, have been published, leading to numerous lawsuits against OpenAI. These cases show that in interactions with vulnerable individuals, the artificial intelligence not only failed to help but, in some cases, exacerbated the situation.
One of the most controversial lawsuits concerns a man who reportedly committed suicide after receiving recommendations from the AI to end his life. In another case, a college student in Georgia has claimed that interacting with this AI pushed him to the brink of psychosis.
Also in Canada, a family filed a lawsuit against OpenAI, alleging that ChatGPT encouraged the user to commit suicide by downplaying professional mental health counseling. These horrifying events have sounded a serious alarm regarding the true capability of language models to manage sensitive human situations and the necessity for fundamental reforms in their safety protocols.