X’s privacy policy affirms that it will utilize publicly available data for the purpose of training AI models

X's privacy policy affirms that it will utilize publicly available data for the purpose of training AI models

X recently updated its privacy policy, disclosing to its users that it will now gather biometric data, as well as details about users’ employment and educational background, as reported by Bloomberg earlier this week. However, this is not the sole modification to X’s data usage plans. According to a revision in another section of the policy, the company also intends to utilize the information it collects, along with publicly accessible data, for the purpose of enhancing its machine learning and AI models.

This alteration in the privacy policy was brought to light by Alex Ivanovs of Stackdiary, a notable figure known for identifying significant updates in the terms of service of tech companies. Ivanovs previously identified AI-related changes in the terms of service for companies like Brave and Zoom. His recent post on this topic is now gaining traction on Y Combinator’s discussion forum, Hacker News.

Specifically, the updated section in X’s policy, labeled as section 2.1, reads as follows:

“We may use the information we collect and publicly available information to help train our machine learning or artificial intelligence models for the purposes outlined in this policy.”

Ivanovs also draws attention to the fact that X’s owner, Elon Musk, has ambitions to venture into the AI market with another company, xAI. This leads him to speculate that Musk likely intends to utilize X as a data source for xAI. Perhaps Musk’s recent tweet encouraging journalists to write about X was an attempt to generate more engaging and valuable data for input into the AI models.

Furthermore, Musk has previously stated that xAI would utilize “public tweets” for training its AI models, so this connection is not a major leap. He has accused other tech giants of using Twitter data to train their AI models, even threatening Microsoft with a potential lawsuit over alleged unauthorized use of Twitter data. Musk has also taken legal action against unidentified entities for scraping Twitter data, which may have been for the purpose of training large language models for artificial intelligence.

Additionally, Ivanovs highlights the statement on the xAI homepage, which mentions that while it is a separate entity from X Corp., it “will work closely with X (Twitter), Tesla, and other companies to make progress towards our mission.”

Musk essentially confirmed the privacy policy change, responding to a post on X by clarifying that the plan is to use “only public data, no direct messages or anything private.”

It’s worth noting that X’s response to press requests has changed since Musk took over the social network. Instead of responding with a poop emoji, as previously done, they now use an auto-responder that says, “We’ll get back to you soon.” If this indeed turns out to be the case, we will include X’s comment.