Twitter began limiting how many posts users can read on its platform, an unusual move that came as owner Elon Musk says he is fighting companies trying to use its data for developing artificial intelligence programs.
Musk on Saturday detailed the temporary changes, which involved limiting unverified accounts to reading just 600 posts a day while verified accounts—those paying a monthly subscription—can read 6,000 posts daily. He later announced multiple increases to the number of allowed posts.
“To address extreme levels of data scraping & system manipulation, we’ve applied the following temporary limits,” Musk tweeted.
Before his tweet, thousands of users complained of problems with the site early Saturday, according to Downdetector, which tracks website outages. Some users said they were getting error messages that said “rate limit exceeded” when trying to access the site. That term was trending on the platform Saturday afternoon.
The move to block Twitter’s own users from reading on the platform flows from an intensifying fight by Twitter to curb AI companies trying to scrape the site for data presumably used to help train their models for free.
The success of OpenAI’s virtual chatbot ChatGPT, including Microsoft’s large investment to get early access to the generative AI technology, has ignited intense interest from startups and investors looking to accelerate their development in the space. That has fueled tensions with social-media companies that have large amounts of data that can help train those systems.
“Almost every company doing AI, from startups to some of the biggest corporations on Earth, was scraping vast amounts of data,” Musk tweetedFriday. “It is rather galling to have to bring large numbers of servers online on an emergency basis just to facilitate some AI startup’s outrageous valuation.”
Twitter users late this past week began noticing new efforts by the company to make such activities harder by requiring web users to be logged into the site view website posts.
“Temporary emergency measure. We were getting data pillaged so much that it was degrading service for normal users!” Musk said Friday of the move. He later added, “Several hundred organizations (maybe more) were scraping Twitter data extremely aggressively, to the point where it was affecting the real user experience.”
Social-media platforms can be a rich trove for the large language models that power AI chatbots because they contain enormous volumes of the types of verbal exchanges those systems are trying to learn. Other social-media platforms also have instituted new policies to deal with the rise of data-hungry AI companies.
Reddit has been facing blowback over its decision to start charging some third-party developers to obtain its data.
“Reddit needs to be a self-sustaining business, and to do that, we can no longer subsidize commercial entities that require large-scale data use,” Steve Huffman, Reddit’s chief executive, said last month in a post on the platform.
Since taking control of Twitter in late October, Musk has been aggressively trying to better control how the company’s user data is used and monetized. Those efforts have included clamping down on use of Twitter’s application programming interface, or API, that lets third parties analyze some of Twitter’s data and build tools with it.
Twitter has increased the price for access to the API as Musk seeks to boost revenue at the company.
In May, Twitter accused Microsoft in a letter of improperly using its data from the API, saying it had been using the data free and was now refusing to pay for it. Musk has accused Microsoft of “illegally using Twitter data” for training.
At the time, Microsoft said it would review the matter and respond when appropriate.
Beyond charging more for its data, Twitter has been attempting to beef up its subscription service that includes verifying users. With Saturday’s move, verified users will be able to read more than unverified ones. Musk’s new rules initially said newly unverified users will be limited to reading just 300 posts a day, before also raising that number multiple times.
“It’s unprecedented as far as I know since the time of” internet bulletin boards, Jason Goldman, a former Twitter executive who has been critical of Musk’s stewardship, said about the limits. “To set limits so low that it renders the site unusable to average users is just the latest in the purely extractive approach the new ownership is taking.”
As the ramification of Musk’s data battle became clear Saturday, some users complained about being limited.
One user tweeted: “Your addiction is about to be broken. We will all be free. See you guys at the park.”
“Touch grass again,” Musk tweeted in reply.
Write to Tim Higgins at email@example.com