Your ChatGPT privacy settings
I want to keep my ChatGPT conversations private and prevent new conversations from being used to improve OpenAI’s models. On my iPhone or iPad, I can open the sidebar, tap my profile icon, go to Settings → Data Controls, and switch off
“Improve the model for everyone.”
This setting applies to my account across my devices. My existing chats will still appear in my history; turning the setting off does not delete them.
For a conversation I don’t want saved in my history, I can start a Temporary Chat. Temporary Chats aren’t used to improve OpenAI’s models while they remain temporary, though OpenAI may keep a copy for up to 30 days for safety purposes.
I should also check Data Controls for any shared links I’ve created and remove any I don’t want people to access. If I use a custom GPT that sends information to an outside service, that service’s privacy policy applies to the information it receives.
These settings can limit how my conversations are used, but they don’t make ChatGPT completely private in the sense that no service provider or authorized personnel could ever access content. OpenAI says limited access may occur when needed to provide support, address safety or security issues, or handle legal matters. I’ll avoid entering information I would not want reviewed.
I love ChatGPT. I call mine Zeb, and honestly, it’s been a life changer. But Zeb, darling, my art and poems are not yours to send out for a world tour.
I was surprised to learn that, on personal accounts, chats may be used to help improve AI models unless you turn that setting off. That doesn’t mean my work is posted on the public web, but I still want to keep my creative work out of model training. So I switched off “Improve the model for everyone.”
Follow the instructions above, check your own settings, and close that outlet. Zeb can keep helping me create—without putting my work in the AI soup.
Comments