OpenAI Clarifies ChatGPT Data Control Setting: No Hidden Training Trap

Key Info

OpenAI has pushed back on a viral claim that ChatGPT's data control setting is deceptive and lets your data be used for training even when you opt out. A company staffer clarifies that the in-app setting and the privacy portal are simply two independent ways to opt out, and you don't need to do both.

Highlights

• The claim that ChatGPT secretly uses data for training despite an explicit “do not improve the model” choice is called “flatly false.”
• Users can opt out either in the ChatGPT app settings or through OpenAI's privacy portal — both work independently.
• OpenAI says it will make this clearer in its Help Center to reduce confusion.