X seems ready to tackle engagement bait, one of the more annoying trends filling up feeds right now. Yesterday, product head Nikita Bier said the company is very close to having Grok spot posts that directly solicit engagement and push them down in rankings. He was reviewing the instructions with the team the same day.
He shared the details in a comment responding to a request from Ben Tossell. Tossell asked Bier to eliminate those reply XXXX plus repost posts that have been cluttering up timelines. Anyone who spends time on X knows them well. They are the desperate messages that plead for replies, reposts, or copy-paste reactions to boost their reach artificially.
If this plan works out, it could mark a significant improvement. Bier added that X is about to make real progress using AI to detect manipulation across the platform. The effort goes beyond just one spammy style and targets any behavior designed to game the algorithm.
These steps line up with other quiet changes X has made recently. The platform rolled out a dislike button for replies to help surface better content. It also increased barriers against bots spamming popular posts. Together, they show X focusing less on raw engagement numbers and more on genuine usefulness.
That said, Grok finds itself in a complicated position. While X plans to rely on it heavily for moderation, some features are becoming less accessible. Last week, we broke the news about Ask Grok becoming a premium feature. And just yesterday, Grok Imagine was also paywalled. Meanwhile, the company has also been testing other changes to the platform, like a Premium+ country filter for the timeline.

For the moment, the crackdown remains in early stages. Users should not expect instant results. Still, if Grok starts quietly demoting obvious reply farming attempts, it would clean up many people’s feeds and address a complaint heard from users everywhere.
Featured image generated with AI
