X head of product Nikita Bier admitted that the platform’s automated Grok content review system is “broken” after an absolute deluge of users complained about harmless posts getting hidden behind adult content filters.
Replying to a thread on Monday, Bier confirmed the issue and stated he has handed the problem over to X’s machine learning team to fix.
The automated filtering system has been going haywire for days. Many accounts are seeing their reach completely tank because regular photos, mildly suggestive images, and even text updates are getting instantly slapped with “sensitive content” warnings.
One user reported getting a warning for sharing a clip of a K-pop star literally just walking down a red carpet.
Another pointed out that their reposts are getting flagged as explicit even though the original creator’s post remains completely untouched on the timeline.
There are many other reports about the issue that you can find on X’s live search feed. So it’s not a niche bug.
This makes it seem like the AI moderation system is overcorrecting and, in turn, unfairly impacting creators who are trying to earn through the monetization program on the platform.
To make matters worse, some users are reporting issues with accessing X’s official content appeal page. That leaves users stuck with restricted posts and hidden media with no way to manually contest the system’s mistakes.
That said, this comes after xAI already started moderating Grok’s output heavily following a high-profile controversy involving explicit deepfakes.
But while that early wave of censorship was confined to what the chatbot itself generated, this new glitch is actively breaking the live feed. It is punishing regular users for entirely safe, everyday activities.
X hasn’t given an official timeline for a fix yet. For now, creators are left hoping the machine learning team tunes the automated reviewer down before more accounts get buried by the algorithm.
Featured image generated with AI


