OpenAI will start monitoring chats in ChatGPT: what will be reported to the police

According to glavcom.ua:

The company OpenAI announced that detected conversations of users on the ChatGPT platform will be checked for disturbing content related to threats of harm to others. In cases where such content is found, it may be handed over to employees for further verification or reported to the police. This decision was made after an increase in cases where chatbots built on artificial intelligence have triggered mental health crises for users, reports 'Glavcom' citing OpenAI.

In its latest blog, OpenAI acknowledged the presence of deficiencies in its systems that may affect users' mental health. The company is now monitoring chats for 'harmful content.' Users expressing intentions to harm others are directed to specialized channels for verification. If the case is considered a threat of serious physical harm, it may be forwarded to law enforcement, as stated in the blog.

However, currently OpenAI does not provide police with information regarding cases related to suicide or self-harm, in order to maintain user confidentiality within the ChatGPT environment.

As a result of OpenAI's announcements, users of the ChatGPT platform can be assured that their conversations will be checked for harmful and disturbing content. This is an important step in protecting users' mental health from negative influences while using artificial intelligence.