X Visibility Filtering: How to Check for a Search Ban
Posts on X stopped showing up? Learn what X's published code says about visibility filtering, how to run a fair check and what X's Terms allow it to limit.
Search your exact username and a recent post while logged out, ask a friend who does not follow you to do the same, and compare your analytics over a week. If posts are missing from search and replies, check for a notice from X and look for the Under the Hood tool. Then appeal any enforcement through X's own routes.
Summary
X's open source algorithm repository says a visibility filtering step decides whether each post is shown, dropped, or shown behind a warning screen.
X's Terms say it may limit the distribution or visibility of content, and its transparency report describes restricting reach as an alternative to removal.
The repository describes a pilot Under the Hood tool showing aggregate stats on visibility labels. X says its availability is still expanding.
A fair check compares a logged out search, a friend's view and your own analytics over several days, not a single test.
One wrong step can cost you the account, so know what you are doing before you start.
Your replies seem to vanish, your name stops showing in search, and views drop to a fraction of what they were. People call this a search ban or shadowban. X does not use those words in the sources we read, but it does publish code and policy that explain how a post can be limited without being removed. Knowing what is written down helps you tell a real limit from a slow week.
This guide reads X's published algorithm repository, the X Terms of Service and X's 2025 transparency report. It sets out what visibility filtering is, how to test your own account without fooling yourself, and what you can and cannot do about it. One wrong step can cost you the account, so know what you are doing before you start.
Search your exact username and a recent post while logged out, ask a friend who does not follow you to do the same, and compare your analytics over a week. If posts are missing from search and replies, check for a notice from X and look for the Under the Hood tool. Then appeal any enforcement through X's own routes.
The README of X's x-algorithm repository describes a visibility filtering component. In the repository's words, it decides whether each post should be shown to a viewer, dropped, or shown behind an interstitial, which is a warning screen you click through. It says it does this using labels attached to posts and accounts, and that those labels come from rules, models and enforcement actions.
The same README separates this from ranking. Ranking decides how high a post appears in a feed. Visibility filtering decides whether it appears at all for a given viewer, and the README says it also takes the viewer's own blocks and mutes into account. That matters for your test: a post hidden from one person may simply be muted by them.
The X Terms of Service, effective 9 October 2026 in the version we read, say X may limit the distribution or visibility of any content on the service. X's 2025 transparency report describes a policy it calls Freedom of Speech, not Freedom of Reach. It says X may restrict the reach of posts to make them less discoverable as an alternative to removal, and that such posts may not be recommended or amplified. The report says X has an appeals process for enforcement.
So a reach limit is a real, documented tool. What the sources do not give is a public switch that tells every user whether one applies to them. That is why a careful check matters more than any single screenshot.
The repository README describes Under the Hood as a pilot transparency tool that shows aggregate statistics on visibility impacting labels, and gives its address as x.com/i/jf/under_the_hood. It says availability is expanding. A September 2026 update listed in the README says the tool now shows legal compliance withholding by country. The README also says a Brazil election visibility filter was removed on 26 October 2026.
Our reading: the tool shows aggregate figures, not a verdict on one account. If the page loads for you, treat it as context. If it does not load, that is not proof of anything, since X calls it a pilot.
Single tests mislead. Search results shift by hour, and your own logged in view is personalised. Run these checks over about a week and note the results.
Open a private browser window, not logged in, and search your exact username. Note whether your profile appears.
Search a distinctive phrase from a recent post of yours in the same window. Note whether the post appears under Latest.
Ask a friend who does not follow you, and has not muted or blocked you, to repeat both searches.
Reply to a large public post and ask that friend whether your reply shows or sits behind a show more prompt.
Compare impressions in your analytics for the same type of post across two or three weeks.
Check your notifications and email for any notice from X about a rule, a label or a limit.
Act on any notice from X first. It is the only official statement about your account.
Use the appeal option in the notice or on X's help pages when you believe a decision was wrong.
Stop the pattern that may have triggered a label, such as repeated identical replies or unmarked sensitive media.
Keep posting normal, original content. A sudden flood of posts to test reach can look like spam.
Keep your own record with dates and screenshots of each check.
The X Privacy Policy notes that your username is public, so a profile that does not appear in search despite a public username is worth recording carefully before you appeal.
Malaysian creators and small sellers using X for promotion can trigger the same patterns as spam accounts, for example by posting the same link to dozens of replies in a short time. If your account mixes Malay and English, ask a test friend who reads both, since a hidden reply in one language is easy to miss. Withholding by country exists in the README, so a post blocked in one country may still show elsewhere.
A clear notice and a simple appeal are things you can handle yourself through X. It gets harder when there is no notice at all, when reach dropped after a hack or a takeover, or when a business account is losing customers while you test. Opening several new accounts to post the same content can be treated as an attempt to evade enforcement and lead to suspension.
If the account matters to your income or reputation, an experienced review of your checks and notices is the safer step. AwareXone does this work. See our X account help page.
A case review helps when your checks point to a limit but X has sent no notice, when an appeal came back with no change, or when the account was misused by someone else. We are a Malaysia-based team, independent of X, and we use only the official routes. We do not ask for your password, and we cannot promise a result. You can read how we work on our trust page.
X's sources we read do not use that word. Its Terms say it may limit the visibility of content, and its transparency report describes restricting reach as an alternative to removal.
What is visibility filtering on X?
X's algorithm repository describes it as the step that decides whether a post is shown, dropped, or shown behind a warning screen, based on labels and the viewer's blocks and mutes.
Can I see if my account is limited?
X's README describes a pilot Under the Hood tool with aggregate stats on visibility labels. It says availability is expanding, so it may not show for every account yet.
How long should I test before deciding?
We suggest about a week of logged out searches, a friend's checks and analytics, because a single result shifts with time and personalisation.
Will deleting posts lift a limit?
The sources do not say. Stopping the pattern that may have caused a label is sensible, and any notice from X is the guide to follow.