The frame will immediately shift to that frame if this enters legal proceedings. The law always views things as you say - only people have agency.

I predict the OpenAI legal team will argue that if a person should be held responsible, it should be the person who originally wrote the content about suicide that their LLM was trained on, and that the LLM is just a mechanism that passes the knowledge through. But if they make this argument, then some of their copyright arguments would be in jeopardy.