‘Incognito mode is now available on ChatGPT thanks to OpenAI

'Incognito mode is now available on ChatGPT thanks to OpenAI

The incognito option for ChatGPT will be introduced by OpenAI and won’t save or use the user’s conversation history for improvement.

OpenAI announced Tuesday that it is offering what one employee referred to as an “incognito mode” for its popular chatbot ChatGPT which does not record users’ interaction histories or use them to further its artificial intelligence.

Additionally, the San Francisco-based business announced plans for a “ChatGPT Business” subscription with more extensive data controls.

The decision was made as concern over how ChatGPT and other chatbots it inspired manage the data from hundreds of millions of users, which is frequently used to “train” or “improve” AI, has grown.

Italy this month suspended ChatGPT due to potential privacy concerns, but said OpenAI might reinstate the service provided certain conditions were met, such as providing users with the ability to object to the processing of their data. Spain and France also started investigating the service.

Chief Technology Officer of OpenAI Mira Murati told Reuters that the business complies with European privacy laws and is striving to reassure regulators.

She said that the new features were the result of months of work to put users “in the driver’s seat” with regard to data collecting rather than Italy’s ChatGPT ban.

With the intention that “it’s completely eyes off and the models are super aligned: they do the things that you want to do,” Murati added, “we’ll be moving more and more in this direction of prioritizing user privacy.”

She said that user data has assisted OpenAI in improving the software’s dependability and reducing political bias, among other problems, but added that the business still has difficulties.

With the product update from Tuesday, users can export their data and turn off “Chat History & Training” in their settings.

When compared to an internet browser’s incognito mode, Nicholas Turley, the OpenAI product officer, said the company would still keep discussions for 30 days to check for misuse before permanently deleting them.

Additionally, discussions won’t be used by default for training AI models under the company’s corporate subscription, which will be accessible in the coming months.

Having invested in OpenAI, Microsoft Corp. already provides ChatGPT to companies. According to Murati, the service would be popular with the cloud provider’s current clients.

Leave a Reply

Your email address will not be published. Required fields are marked *