OpenAI reveals that data from free and Plus users is used to train its models until changed

OpenAI has revealed that content shared by free and Plus users is used to train its models; until turned off, it is on by default and keeps using your data. We have seen this one before, remember? Yes, Microsoft has its newly unveiled Recall feature in Copilot+ PCs.

After public bashing and proving how it can be a privacy hazard, Microsoft finally turned it off by default. A rare win for users but an important one.

What I also noticed is that this setting of theirs is hidden under a very clever name: “Improve model for everyone.”

However, OpenAI says that privacy is a top concern. Users have the option to opt out of contributing their data to model training through their settings. OpenAI also revealed that certain interactions, like “Temporary Chats” in ChatGPT, are excluded from training data by default. Business plans, such as API, Enterprise, and Team, also have their data excluded by default.

Here’s how to do it:

• Click on your profile picture or initials in the top-right corner of the screen.
• Select “Settings” from the dropdown menu.
• Within the settings menu, look for an option related to “Data Control.”
• Find the toggle switch or checkbox for “Improve model for everyone.”
• Turn off this setting.

OpenAI emphasizes that its goal is to use data to improve the general understanding of language, not to create individual profiles of users. They guarantee that public data is not used for targeted advertising or user profiling.

The company also said to use privacy-preserving techniques. Their models are designed to minimize the amount of personal information used during training.

More here.