X Officially Bans Using Its Content to Train AI

The social media platform X (formerly known as Twitter) has introduced new rules aimed at restricting the use of its data by artificial intelligence companies. On June 4, the company updated its Developer Agreement, specifically the section titled "Reverse Engineering and Other Restrictions", by adding the clause:

“You may not, and may not attempt to (or permit others to)… use X's API or content for training a machine learning or artificial intelligence model, including foundation or advanced models.”

This move reflects a significant policy shift following the acquisition of X by xAI — an AI firm launched by Elon Musk. The strategic change is intended to prevent competitors from freely accessing X’s vast user-generated content to train their own large language models (LLMs) without a formal licensing agreement.

In 2023, X altered its privacy policy to allow internal use of public user data for AI training. Later that year, the platform permitted external developers to do the same — a privilege now officially revoked. The new rules suggest tighter control over how X’s data ecosystem is exploited in the rapidly evolving AI landscape.

The company’s new direction reflects Musk’s vision of integrating social data into xAI’s proprietary systems, while simultaneously blocking third-party AI labs from capitalizing on it. This raises important questions around data ownership, open access, and AI ethics.