In short, I had disabled the setting “Improve model for Everyone” a while ago. At some point (possibly when updating the iOS GPT app?) it became enabled again without my awareness
I had asked GPT to perform a literary analysis on unpublished written works of mine before realizing the setting was re-enabled. This was not a temporary chat. Is GPT going to use this attachment for llm training down the road? I asked GPT directly and it said no today, but yes last night
When the “Improve the model for everyone” setting is enabled in ChatGPT, your conversations and attached files MAY BE used to train and enhance OpenAI’s models.
I suspect it probably will try to train on the uploaded pdf and I really, REALLY do not want that happen because these are novel literary works I have spent years writing and it would be extraordinarily disappointing to see someone take those styles and themes as their own, before I even have a chance to publish as an author. I have submitted privacy requests and contacted support with the details, but given the sheer volume of inquiries they probably get vs their team size I have my doubts they will address my concerns in time before the pdf is trained upon. Is there anything else I can do to prevent this?
FYI I’m. I am not against AI art but I didn’t even get a chance to publish or capitalize on this before it’s getting sent to the AI machine!! Literally years worth of work
As you submitted a request about privacy, this is the correct way you made. But, I never sent that kind of submition and I don’t know when they reply in hours or days.
BTW, for future reference about safety of your data, you may visit following links from OpenAI help pages: