X users are starting to discover that their posts and interactions are being utilized to enhance Grok, a generative AI chatbot. This practice was initiated without prior notification.
In X’s settings, a newly added “Data sharing” toggle suggests that any data created or interacted with on the platform may contribute to “training and fine-tuning” Grok. Previously, Grok’s training was limited to a disparate set of texts available publicly online, and it did not approach the sophistication or breadth of data sets used by competitors like ChatGPT or Co-Pilot.
X maintains that allowing Grok to access a live stream of public X posts enhances its ability to deliver timely and diverse responses to user queries. In theory, Grok aims to better understand human language and communication, deliver more relevant and engaging replies, and develop a balanced viewpoint without political bias, even attempting to cultivate a sense of humor.
However, the manner in which X has rolled out this feature raises significant privacy concerns. The capability to contribute data to Grok’s training is set to on by default and the option to opt-out is somewhat obscured, found only within the web version of X. This implies that many users, particularly those accessing X via mobile, are unwittingly training the AI with their data.
Users concerned about their digital footprint and data privacy must navigate to the “Privacy & Safety” settings on X’s web version to disable the default setting under “Data sharing and personalization” for Grok.
- On X web select “Privacy & Safety”
- Under “Data sharing and personalization,” select “Grok”
- Turn off the option “Allow your posts as well as your interactions, inputs, and results with Grok to be used for training and fine-tuning.”
