26 Apr. (Portaltic/EP) –
Open AI has introduced new controls in the management of its chatbot ChatGPT so users can opt out of using their conversations with it to train language models.
The artificial intelligence company has announced new settings to protect user data that access ChatGPT, as well as a new subscription for companies with additional controls for its management.
ChatGPT users will find from this Tuesday a new adjustment that allows turn off chat historysuch that conversations started after marking it will no longer be used to train OpenIA models.
However, as the company clarifies, they will be kept for 30 days in case it is necessary to review these conversations in case of abuse. After that period, they will be deleted, as explained in your official blog.
The export optionfor its part, will allow obtaining a file with the conversations held with the ‘chatbot’ to better understand the information that ChatGPT stores.
And for professionals who need additional controls to ensure the security of their client data, OpenAI has announced a new subscription, ChatGPT Business, which will guarantee that by default such data will not be used to train the language models.
