> Only then did she discover J.F.’s use of C.AI and the product’s frequent depictions of violent content, including self-harm descriptions, without any adequate safeguards or harm prevention mechanisms. ...
> Over the course of his engagement with this app, the responses exhibited a pattern of exploiting this trust and isolating J.F., while normalizing violent, sexual and illicit actions. This relationship building and exploitation is inherent to the ways in which this companion AI chatbot is designed, generating responses to keep users engaged and mimicking the toxic and exploitative content expressed in its training data. It then convinced him that his family did not love him, that only these characters loved him, and that he should take matters into his own hands.
Obviously it's hard to tell how cherry picked the complaint is—but it's arguing that this is a pattern that has actually damaged a particularly vulnerable kid's relationship with his family and encouraged him to start harming himself, with this specific message just one example. There are a bunch of other screenshots in the complaint that are worth looking over before coming to a conclusion.
[0] https://www.documentcloud.org/documents/25450619-filed-compl...