Elon Musk’s Facebook X platform is under pressure from data regulators after it emerged that users are consenting to their posts being used to build artificial intelligence systems through a default setting in the app.
Data watchdogs in the UK and Ireland said they had contacted X over the apparent attempt to obtain users’ consent for data collection without their knowledge.
A user X highlighted the issue on Fridaypointing to a setting in the app that was turned on by default and allowed the account holder’s posts to be used to train Grok, an artificial intelligence chatbot created by Musk’s company xAI.
Low UK GDPRwhich is based on the EU data regulation of the same name, companies cannot use “pre-checked boxes” or “any other default method of consent.”
The setting, which comes with a box already checked, states that it “allows your posts, as well as your interactions, inputs, and results with Grok to be used for training and tuning.” According to user X, the setting can only be used Disabled in the web version of X.
Data regulators immediately raised concerns about the default settings. In the UK, the Information Commissioner’s Office (ICO) said it was “making enquiries” with X.
“Platforms seeking to use their users’ data to train their core AI models must be transparent about their activities,” an ICO spokesperson said.
“They should take steps to proactively notify users well in advance of the use of data for these purposes, and provide them with sufficient time and simple processes to object to their data being used in this way.”
The Republic of Ireland’s Data Protection Commission (DPC), the main regulator for X in the European Union, said it had already been speaking to Musk’s company about data collection and AI models this week and was “surprised” to learn of the default settings.
“The DPC has been in contact with X on this matter for several months, with our last interaction occurring yesterday, so we are surprised by today’s developments. We have continued to be in contact with X today and are awaiting a response. We look forward to further contact early next week,” said Graham Doyle, Deputy Commissioner of the DPC.
Big data language models are the technology that underpins chatbots like ChatGPT and Grok and are fed with vast amounts of data scraped from the internet to detect patterns in language and develop a statistical understanding of it. This ultimately enables chatbots to generate compelling responses to queries.
However, this approach has met with opposition in multiple quarters, including from news publishers, authors (who argue the process violates copyright laws) and regulators.
This month, Facebook and Instagram owner Meta said it would not launch an advanced version of its artificial intelligence model in the EU, blaming the decision on “unpredictable” behaviour by regulators.
X has been contacted for comment.