Programmatic replies on X now face strict limits, as part of ongoing efforts to reduce spam on the platform. X has rolled out this change on February 23. The platform will now restrict replies sent through the POST/2/tweets endpoint. Developers can only post these replies when the original author mentions their account or quotes their post.
This update is a major step in tackling automated reply spam. Low-quality bots and AI replies flooded conversations with instant replies, which were repetitive and plain bad. Users saw fake Elon Musk accounts, crypto spams, or random automated chatbot replies. These replies used to appear seconds after popular posts dropped, overall lowering the quality of content and discussions on the platform.
Since these restrictions are already in effect, it should help keep the replies section a lot cleaner. The rule applies to Free, Basic, Pro, and Pay-Per-Use plans. Enterprise customers and Public Utility apps are exempt from this rule. Automated main posts (not replies) are allowed to continue operating with no changes to the rules.
Earlier, developers used to run scripts that would automatically spam low-quality AI replies instantly to popular posts. This will now completely fail. Spam with API v2 was the easiest entry point that allowed low-cost, high-volume spam replies. The ‘Reply Guy’ monetization strategy will no longer work if you don’t authentically type your replies on your own.
Manual replies obviously aren’t restricted, and these new rules only apply to programmatic automation.
Nikita Bier, the head of product at X, described the plan clearly. He called it “Operation Kill the bots.” The first step is to ‘close the front door,’ and automated replies are now only possible if the original author invites them to participate via a mention or a quote.
The efforts from X to reduce spam on the platform are heavily appreciated. Recently, they’ve also been testing a label to mark AI-generated content as ‘Made with AI,’ which you can read about here.

