X’s new policy prevents companies from using posts to ‘fine-tune or train’ AI models

X has updated its developer statement to adhd a caller regularisation connected utilizing posts connected the level to bid AI. The updated policy, spotted earlier by TechCrunch, says developers can’t usage contented from X oregon its API to “fine-tune oregon bid a instauration oregon frontier model.”

The argumentation alteration could acceptable up X to marque AI grooming deals with third-party companies, akin to the woody Reddit struck with Google. Reddit, which has a akin argumentation to block AI crawlers, sued Anthropic connected Wednesday implicit claims the company’s AI crawlers accessed the tract much than 100,000 times since July 2024. Elon Musk’s AI company, xAI, acquired X for $33 billion connected insubstantial successful March. 

Though X’s developer statement present bars companies from grooming AI connected its content, its privateness argumentation inactive states that third-party “collaborators” tin bid AI models connected the site’s information unless users opt out. X besides feeds idiosyncratic information into its AI model, Grok, for training.