Content
summary Summary

OpenAI responds to privacy criticism from international companies and government agencies by addressing two of the biggest criticisms with new data control settings.

Ad

First, ChatGPT users can now turn off chat history. Turning it off also ensures that data entered in chat will not be used to optimize future OpenAI AI models.

However, the chat history will no longer appear in the ChatGPT sidebar, so you will lose an important feature. Generated chats with the chat history turned off cannot be restored.

Regardless of the disabled chat history, OpenAI will continue to store the chats for 30 days, but will only evaluate them if there is a suspicion of abuse. After 30 days, the chats will be permanently deleted.

Ad
Ad

OpenAI will continue to store chats and use them to optimize the model for those who leave chat history enabled.

New ChatGPT data controls

Disabling chat history is available in the ChatGPT settings under "Data Controls" for both paid and free accounts. Chat history is enabled by default and must be disabled here to prevent your data from being used for continuous AI training.

In addition to disabling chat history, the new data control settings also offer an option to export data and an option to delete your account. Deleting an account will delete all data associated with that account and cannot be undone. This also applies to accounts that use the API. The export function sends all ChatGPT chats and associated data to the account address via email.

More privacy-friendly ChatGPT Business will be available in the coming months

For businesses, OpenAI intends to offer a new service called ChatGPT Business "in the coming months". By default, user data will not be used to train models, but chat history will be kept. ChatGPT Business for professional users and enterprises uses the same privacy policies as the GPT API.

If you would rather not wait for ChatGPT Business, you can use this form to unsubscribe from model training immediately. After submitting the form, new conversations will not be used for model training.

Recommendation

The new privacy settings are OpenAI's response to increasing criticism from European data protection authorities, as well as large companies banning their employees from entering corporate data into ChatGPT. The coming weeks will show whether these measures are enough to quell the criticism.

Ad
Ad
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.
Support our independent, free-access reporting. Any contribution helps and secures our future. Support now:
Bank transfer
Summary
  • OpenAI responds to privacy criticism from international companies and authorities by offering new privacy options.
  • From now on, you can opt out of training the AI model with your data - in exchange, your chat history will be disabled.
  • For businesses and professional users, OpenAI plans to offer ChatGPT Business "in the coming months," which will disable the training opt-out by default while still retaining chat history.
Online journalist Matthias is the co-founder and publisher of THE DECODER. He believes that artificial intelligence will fundamentally change the relationship between humans and computers.
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.