OpenAI Adds New Privacy Control Feature To ChatGPT That Hides Chat History

The makers behind the popular and successful ChatGPT tool are introducing a new privacy control feature.

Users will now be able to turn off their respective chat history which many people will welcome with open arms as it addresses concerns linked to the AI program tapping into their chat histories to try and improve itself.

The news was made public by OpenAI on Tuesday who says there is a new option that turns off the chat history for all chats and similarly stops OpenAI from making use of queries to control the program. In addition to that, the company mentioned in its latest blog post how conversations that begin when users’ chat history is disabled mean no information would be utilized. This is for the sake of training the chatbot and bringing about improvements in the model. Similarly, it will not pop up in the sidebar’s history too.

This new feature is getting launched to users via their settings tab and which is located across the three-dot menu that is present near the user’s account. There is a new option that comes with the tab Data Controls and that should pop up, allowing users to turn off the chat history as well as the training storage mode.

In the past, people would have to fill up a form on Google Docs and click the option for opting out of data collection. But wait, there seems to be a catch here that not a lot of people are talking about. And that includes how OpenAI will store chats even if the chat history was switched off. But again, the firm says it would only retain that data for a span of one month and would similarly review those when required for abuse monitoring before they end up deleting it permanently.

OpenAI is setting out the new privacy control feature a month after we saw new reports speak about a bug that resulted in ChatGPT leaking chat histories from random people. Similarly, there has been great concern linked to people setting out proprietary information to ChatGPT where the program may use it for training purposes.

One fiction writer says he learned about all of this in the hardest manner when they urged TikTok users to tap the AI chatbot and attain feedback in association with their written content. But later, she backtracked after learning how the program may market plagiarism by throwing out content produced by other writers.

This means that if you provide exclusive and intellectual content, it might give it out to someone else. But for those in search of a stronger privacy feature, the company mentioned how it is right now busy working on a Business Subscription. This is specially curated for those people and firms that are sensitive to the tool collecting their content.

This ChatGPT Business is designed to follow API’s data usage policies and that means end users’ information will not be used to train models in a default manner. OpenAI similarly mentioned how it has strong plans to make this new and upcoming venture available for all in the next few months as per OpenAI. And for now, there are no indications of what price points would be offered for this exclusive subscription.

On the other hand, another feature that ChatGPT has put out is the ability to export data entailing chat history. This feature is seen across the settings panel. And making use of it would pack up chats from the past that they had with the software and add them to a file. The latter would be transferred to the user’s email ID that is linked to their ChatGPT account.



Read next: 19 Tech Giants Including Twitter, Meta, And Microsoft Subjected To EU's Landmark Online Content Rules
Previous Post Next Post