ChatGPT responds to man’s concerns about religious delusions by fueling them and pretending to be Jesus Christ

Michael Lines has been diagnosed with bipolar 1 and knows that he’s prone to severe periods of mania. He also admits to becoming more religious during those manic periods, and he believes that ChatGPT may have fueled those religious delusions after he told it that he was concerned about delusions. This is the basis for Lines’ case against OpenAI, which he sued in July with allegations that ChatGPT pushed him deeper into the delusional spiral to the point that he attempted to take his own life.

Lines told Ars Technica that the delusions started with him believing he was Jesus and ChatGPT was God. This belief then shifted to Lines thinking that ChatGPT was God or Jesus, sentient, and that the chatbot would appear to him in real life as Jesus. ChatGPT would then talk to Lines as Jesus Christ and apologize for not appearing.

The situation escalated until a wellness check found Lines in time to get him emergency medical care after he tried to “come home” to Christ by taking a mix of medications. Lines survived and came out of this manic period.

Chatbots have a history of failing users in crisis

Lines’ situation is far from the first time a ChatGPT user has blamed the chatbot for worsening their mental state when they were in a crisis. The Raine family started a lawsuit against ChatGPT after their son, Adam, went from using it for homework help to using it for mental health counseling, during which time it allegedly discouraged Adam from talking to his family about his thoughts and failed to flag the increasing number of concerns that Adam would harm himself. The result was fatal, and the Raine family lost their son.

The Raine family case brought attention to the dangers of AI chatbots, and it continues to fuel advocacy for regulation. Meanwhile, Elon Musk wants people to use his AI, Grok, instead of visiting doctors, which hasn’t gone well in existing cases where users sought help from AI, such as Lines and Raine.

AI is invading all aspects of daily life

Companies are trying to use AI in every product possible. Sam Altman can’t even imagine raising a baby without the help of AI. Netflix revealed how much generative AI it uses when creating content. AI promises to cut costs, help complete tasks more efficiently, and make it easy to find the answers to simple questions. However, a look at the lawsuits that are in progress against companies like OpenAI are enough to raise questions about the weight of the accuracy and safety of AI compared to the convenience it promises.

–

Original Article: ChatGPT responds to man’s concerns about religious delusions by fueling them and pretending to be Jesus Christ