Reddit is closing several of the easiest ways for outside tools to collect its public conversations, as the social news platform steps up efforts to control large-scale scraping for AI and automated access to user posts.
Publicly known as the front page of the internet, the company says it will stop supporting RSS feeds on November 13, 2026, begin cutting off unregistered public API users from January 12, 2027, and end remaining public Data API access in March 2027.
It is also tightening access to Old Reddit, its legacy desktop site, by limiting it to moderators and logged-in users who have used the interface within the previous six months.
Why Reddit Is Cutting Open Access
Reddit says the changes are aimed at large-scale scraping and automated abuse. In its announcement, the company said RSS, a simple web feed that lets people and software automatically receive new posts without repeatedly visiting a website, had become a common route for this activity.
Once RSS support ends, Reddit says there will be no replacement for people using feeds outside communities they moderate. And so, moderators who rely on RSS alerts are being directed to tools on Reddit’s Developer Platform, including Discord Relay.
The public API is also moving behind tighter controls, and more than 14,000 apps and bots have already registered for migration, according to Reddit. New requests for public API access will stop on October 31, while apps that have not registered or responded to Reddit’s outreach will begin losing access in January.
AI Data Has Become More Valuable
The restrictions come as Reddit’s huge archive of user discussions has become increasingly valuable to artificial intelligence (AI) companies looking for human-written data.
AI developers need large amounts of text to train and improve their models, and Reddit offers years of conversations covering everything from technology and finance to hobbies and personal experiences. As such, this information can be collected using automated software known as web crawlers or scraping bots, which visit pages and copy large amounts of content much faster than a person could.
Reddit’s own public tools have also made some of its data easier for software to access, especially as RSS feeds allow users and automated tools to receive new posts in a standard format, while Reddit’s public API lets approved software request posts, comments and other information directly from the platform.
But as demand for this type of data has grown, Reddit has increasingly sought to control who can access it and under what terms. The company already has licensing relationships with Google and OpenAI, where its 2024 agreement with Google was reportedly worth about $60 million a year and allowed Google to use Reddit content for AI training. Reddit also later signed a separate partnership with OpenAI that gives the company structured, real-time access to Reddit content.
At the same time, Reddit has taken action against companies it says collected its data without permission. In June 2025, it sued Anthropic, accusing the Claude developer of accessing or attempting to access Reddit more than 100,000 times after Anthropic had said it blocked its bots from the platform. However, Anthropic denied Reddit’s claims and said it would defend itself.
What Changes for the Open Web
These latest restrictions affect more than AI companies. Researchers, social listening services, independent developers and users who rely on RSS or the public API will also have fewer ways to collect Reddit posts automatically.
Reddit had already changed its robots.txt file in 2024 and said it would rate-limit or block unknown bots and crawlers. But its rules now explicitly prohibit scraping without an authorised agreement and bar developers from using Reddit data to train AI models without permission.
And these measures show why the latest restrictions matter, as Reddit is making it harder for outside tools to collect its content freely while directing more automated access through systems and agreements it can approve and monitor.
And while the company says it is offering migration support, including a $1 million programme for eligible developers moving apps to its Developer Platform, it still remains clear that large-scale automated access is increasingly being moved into systems the company can approve, monitor, or license.
