X Secretly Begins Using User Interactions, Inputs, and More to Train Grok AI by Default

Users have begun to take notice that X (formerly Twitter) has begun to use interactions, inputs, and post results to train Grok’s AI model. This setting is on by default but an option to turn it off is available.

Recently, users of the social media platform have begun to take notice of a new suspicious setting related to Grok and data sharing. First seen by @ichigane, this setting was seemingly slipped by X users without any public announcement. Hidden within the Privacy and Safety settings is a new page dedicated to the X AI platform Grok.

Under ‘Data Sharing’ is a setting to “Allow your posts as well as your interactions, inputs, and results with Grok to be used for training and fine-tuning.” The setting goes on to state that interactions, inputs and results “may also be shared with our service provider xAI for these purposes.”

Thankfully on the web version of X, users can simply turn this setting off. There is also an option to ‘Delete conversation history’ which is likely the extent of wiping away any data being used for Grok. To do so, navigate to ‘Settings’ and ‘Privacy and Safety’. You’ll then have to scroll down to find the ‘Grok’ menu. It’s worth noting that the mobile version of X does allow you to turn off this setting.

Since launching the Grok AI Chatbot earlier this year, X has continuously attempted to build out its feature set. Premium paying users can not only use Grok as a basic chatbot. On top of that, Grok offers the ability to summarize news and trending events. Recently, it also launched a ‘More About This Account’ feature, which uses AI to summarize a user in a few sentences.