Maybe I'm misunderstanding but it seems a little dubious to me that OpenAI could just rely on ToS to prevent direct copying of the source material. Shouldn't there be some onus on them to not spit out portions of stories verbatim? Rather than just blaming it on people for writing 'banned' prompts?
