The New York Times has updated its terms of service, prohibiting AI companies from using its content for model algorithm training. The new terms explicitly state that articles, videos, images, and metadata from the Times cannot be used to train any AI models without permission. This restrictive usage clause has sparked controversy over the use of scraped data for training AI models, a debate that has yet to be fully resolved in U.S. courts. This move could potentially lead to new AI litigation or regulations. Microsoft has also added similar restrictions in its terms and conditions, prohibiting the use of its AI products to create, train, or improve other AI services, and banning the scraping or extraction of data from its AI tools. Content that has already been scraped has become part of GPT-4, but whether it respects the wishes of content owners remains to be seen.