Free trialTry Starter or Influencer free for 2 days.Start free trial →
Account Safety

Twitter Shadowban: How to Tell If You're Shadowbanned and How to Fix It

Short answer

A Twitter shadowban is X limiting how far your posts travel without telling you. X calls it visibility filtering and applies it through labels on posts or accounts. Since August 2026 you can see those labels yourself on X's Under the Hood page (x.com/i/under_the_hood), which lists the labels applied to your account and posts in the previous month. If that report is clean, run the logged-out search and reply tests before you assume you're shadowbanned.

Key takeaways

  • X doesn't use the word shadowban. Its systems attach labels to posts and accounts, and each label has a defined effect, most often hiding the post from recommendations to non-followers.
  • X's Under the Hood page shows the labels applied to your account and posts in the prior calendar month, if your account is at least a year old and posted 10 or more times.
  • The label list is public in X's open-source repository, so you can read exactly what SPAM_HIGH_RECALL or Compromised means and what it does to your reach.
  • Third-party shadowban checkers only probe public surfaces (search, suggestions, replies) and infer a result, which is why two tools can disagree about the same account.
  • Most reach drops are not shadowbans. A clean report plus visible posts in logged-out search means the problem is distribution, not punishment.
  • Fixes follow the label: remove or edit flagged posts, stop the pattern that triggered it, secure the account if it was flagged as compromised, and re-check the next monthly report.

Twigor sells growth plans for X accounts. We clarify where our product fits and where it does not, and every claim about X's rules links to the source.

The post did fine last week. This week the same kind of post gets eleven views, none of them from strangers, and a friend says she searched your name and you didn’t come up. Nothing in your notifications. No email from X. Just a quiet drop.

For years the only way to investigate that feeling was a third-party checker and a lot of guessing. That changed in August 2026, when X published the code that decides whether a post gets filtered and added a page where you can see the labels sitting on your own account. So this guide starts there, then covers the manual tests that still matter, and then what to do about what you find.

What Is a Twitter Shadowban?

A shadowban is X limiting the reach of your posts or account without telling you, and X’s own name for it is visibility filtering. The word “shadowban” never appears in X’s rules. What the rules describe is a set of enforcement options short of suspension: making a post less discoverable, leaving it out of search results and trends, downranking it in replies, or removing it from recommendations entirely.

X laid out the philosophy in 2023 under the phrase “freedom of speech, not reach.” A post that breaks a rule but doesn’t warrant removal stays on the platform and gets less distribution. In practice that is what people feel as a shadowban: you can still post, your followers can still find you, and almost nobody else does.

Under the hood this works through labels. X’s automated systems, and sometimes a user report, attach a label to a post or to an account. Each label has a fixed effect. Since August 2026 the list of labels X is willing to show you, with the plain-English meaning of each one, is public in the open-source x-algorithm repository. That turns a vague fear into something you can look up.

How Do I Check If I’m Shadowbanned on X?

Start with X’s own report, then confirm with two logged-out tests. Don’t start with a third-party checker. We’ll get to why.

Open your Under the Hood report

Log in and go to x.com/i/under_the_hood (newer links use x.com/i/jf/under_the_hood). If your account is eligible, you’ll see two sections: account labels and post labels, covering the previous calendar month. Each entry says what the label means and what it did to visibility, and you can download the whole thing as JSON.

Eligibility, as of November 2026: the account has to be at least a year old and has to have posted 10 or more times in that month. X launched it as a pilot for a randomized group, then widened it, added legal-compliance limits in September, and in late October started listing which specific posts carried a label. If you see nothing at all, you may not be in the rollout yet, or you didn’t post enough last month to generate a report.

A report with zero labels is the single most useful piece of evidence you can get. It means X’s systems didn’t flag your account or posts last month, and whatever happened to your reach happened somewhere else.

Check your Account status page

In Settings, under Your account, X shows an Account status screen. Explicit enforcement actions, the kind with a reason and an appeal button, show up here. Under the Hood covers automated labels; Account status covers the ones a human or a policy decision put there. Look at both.

Run the logged-out search test

Open a private browser window so you’re logged out. Search for from:yourhandle and switch to the Latest tab. Your recent posts should be there. If nothing appears, or only posts from months ago, your posts are being left out of search. Then type your handle, without the @, into the search box and see whether your account appears in the suggestions dropdown. Missing from suggestions is the mildest restriction and the most common one checkers flag.

If you’d rather not do this by hand, our free Twitter Shadowban Test runs the same searches, plus the reply check below, for any public account. It shows the post each result is based on, and it answers “Can’t tell” when there isn’t enough to go on. It reads the same public surfaces as every other checker, so use it as a quick second opinion after your Under the Hood report.

Look at your replies from an account that doesn’t follow you

Find a reply you left under a large account’s post. Open that post from a second account, or ask someone who doesn’t follow you to open it, and see whether your reply is visible without clicking “Show more replies.” A reply that’s consistently tucked behind that button is being downranked. One or two is normal. All of them is a pattern.

Compare impressions, not likes

Open your analytics and look at impressions per post over the last four weeks, split by followers and non-followers if your plan shows it. A shadowban almost always shows up as non-follower impressions falling to nearly zero while follower impressions hold. If both fell together, you probably posted less or posted differently, and that isn’t filtering.

What Do the Under the Hood Labels Mean?

Each label in the report has a stated cause and a stated effect, and most of them do the same thing: hide the post from recommendations to people who don’t follow you. The wording below is X’s own, from the public label file.

LabelWhat it meansWhat it does
SPAM_HIGH_RECALL (post)Automated systems think the post may contain spamHidden from recommendations to non-followers
SpamHighRecall (account)Automated systems think the account is likely to post spamAll the account’s posts hidden from recommendations to non-followers
MALICIOUS_URL (post)Likely links to a malicious siteHidden from recommendations to non-followers
NSFW_HIGH_RECALL / NSFW_TEXT (post)May contain adult content or explicit languageHidden from non-followers, underage users, users with no stated age, and logged-out visitors
NSFW_HIGH_PRECISION (post)Likely adult content, by automation or a reportShown behind a content warning, plus the hiding above
FOSNR_ABUSE / FOSNR_HATEFUL_CONDUCT / FOSNR_VIOLENT_SPEECH (post)Found to violate the abuse, hateful conduct or violent speech policy, usually after a reportDiscoverable only on your profile, with a visible “limited visibility” notice
Compromised (account)Account looks taken overPosts hidden from non-followers until you reset your password
ReadOnly (account)A post that violates the terms must be removedAccount can only read until it’s removed; posts hidden from non-followers
SPAM / PDNA / BOUNCE (post)Found to violate authenticity or terms rulesPost not shown on X at all, or not shown pending review or deletion

Two things worth noticing. First, “hidden from recommendations to non-followers” is the standard effect, and it describes exactly the symptom people report: followers still see me, nobody new does. Second, the labels that come from a user report (the FOSNR family) are not silent. X says all users can see a notice that the post has limited visibility. So if you’ve been reported and limited for abuse, you’d know. The quiet ones are the automated spam and adult-content labels.

The file calls itself a public allowlist, meaning these are the labels X will display in your report. X doesn’t claim they’re the only labels that exist. Treat the report as strong evidence, not a complete audit.

What Gets an Account Shadowbanned?

Labels are triggered by patterns that look like spam, adult content, malicious links, or a compromised account, either to an automated classifier or to someone who reported you. The label names tell you the categories. The patterns behind them, from what X has published and what we see when people come to us after a drop, are mostly these:

  • Repetitive posting: the same text, the same link, or the same reply pasted under dozens of posts in a short window.
  • Follow churn: following hundreds of accounts and unfollowing them days later, which X’s platform manipulation policy names as a form of spam.
  • Links to domains X’s systems have classified as malicious, including shortened links that redirect somewhere flagged.
  • Adult or violent media in posts, in your avatar, or in your banner image. Several of the account labels in the file are specifically about the avatar and banner.
  • Login patterns that look like an account takeover: new devices, new countries, sudden bursts of activity. That earns the Compromised label and a forced password reset.
  • A cluster of user reports on a single post, which can produce one of the FOSNR labels.

One pattern that doesn’t earn a label, as far as the public file shows: posting about a controversial topic. The labels are about spam, adult content, abuse and legality, not about opinions.

How Do I Fix a Shadowban on X?

Match the fix to the label, then stop the behavior long enough for the next monthly report to come back clean. There is no “remove shadowban” button, and there’s no customer-support shortcut. There is a sequence that works.

Read the label and act on what it says

A post label applies to that post. If your report shows SPAM_HIGH_RECALL on three posts, delete or edit those three. An account label applies to everything, so the fix is the behavior, not any single post. If the label is Compromised, X has already told you what to do: reset your password and turn on two-factor authentication.

Cut the pattern for a full week

Stop the thing that looks like spam. No bulk following or unfollowing, no repeated links, no copy-pasted replies. Post normally, a few times a day at most, with original text. Automated labels are applied by classifiers watching recent behavior, so the fastest way to stop looking like spam is to stop acting like it.

Appeal through Account status if there’s a formal action

If the Account status screen shows an enforcement action, it also shows an appeal path. Use it, keep it short, and say specifically what you’ve removed or changed. Automated labels in Under the Hood don’t have an appeal button; those lift on their own when the signal stops.

Check the next report

Under the Hood covers whole calendar months. If you clean things up mid-November, December’s report is the one that tells you whether it worked. In the meantime, run the logged-out search test weekly, by hand or with the shadowban test. Search visibility usually returns before the report confirms it.

Don’t try to post your way out

Doubling your posting volume while labeled makes the spam classifier more confident, not less. And buying followers or engagement to “restart” reach pushes you toward the exact labels you’re trying to lose.

Is Low Reach Always a Shadowban?

No, and the ranking code explains why: distribution to strangers depends on predicted engagement, and a clean account can still have almost none. The same open-source release that published the labels also published the For You ranking pipeline. For every viewer, X predicts how likely they are to like, reply, repost, dwell on, or report each candidate post, and ranks by a weighted sum of those predictions. A post with a link, a post that nobody replies to, a post from an account with no engagement history with that viewer: all of these score low without any label involved.

So the honest diagnosis after a clean report is uncomfortable but useful. Your posts aren’t being hidden. They’re being out-ranked. That’s a distribution problem, and distribution problems have different fixes: posts people reply to, replies under larger accounts in your niche, and getting your posts in front of the right people often enough that the model learns who wants them. That last part is what Twigor’s promotion plans are built for, and it only works on an account that isn’t labeled, which is why we tell people to run the checks above before they spend anything on growth.

If your reach drop came with a falling follower count, read Why Am I Losing Followers on X? as well. The two problems get blamed on each other constantly, and they almost never have the same cause.

Conclusion

The shadowban stopped being a mystery in August 2026. Open Under the Hood, read the labels, and you know whether X’s systems flagged you and what it did. If the report is clean and your posts show up in a logged-out search, you aren’t shadowbanned, and the next question is why strangers aren’t engaging. If the report isn’t clean, the label tells you what to fix. Either way, you’re working from X’s own data now, not from a checker’s guess.

Frequently asked questions

How do I know if I'm shadowbanned on X?

Open x.com/i/under_the_hood while logged in. If your account qualifies, it shows the labels applied to your account and posts last month and what each one does. Then log out and search from:yourhandle. If the report is clean and your recent posts show up in search, you're not shadowbanned in any meaningful sense.

Is there an official shadowban checker for X?

Yes, as of August 2026. X's Under the Hood page lists visibility-limiting labels on your account and posts for the previous calendar month. It requires an account at least one year old with 10 or more posts in that month, and it started as a pilot, so not everyone sees it yet.

What does SPAM_HIGH_RECALL mean?

It's a post label from X's automated systems meaning the post may contain spam. Its effect, in X's own wording, is that the post is hidden from recommendations to non-followers. Your followers still see it; strangers in the For You feed don't.

How long does a shadowban last on X?

There's no published duration. Post labels stay with the post, so a flagged post stays limited until it's deleted or the label is reviewed. Account labels like SpamHighRecall are applied by automated systems and tend to lift once the behavior that triggered them stops, but X doesn't publish a timeline.

Can a shadowban checker be wrong?

Often. Checkers don't read X's labels. They test whether your posts appear in search, whether your handle autocompletes, and whether your replies are collapsed, then guess. A new account, a protected account, or a post with a link can trip those tests without any label on the account.

Does posting links get you shadowbanned?

A link alone doesn't earn a label, but a link X's systems classify as malicious does (MALICIOUS_URL hides the post from recommendations to non-followers). Separately, link posts tend to get less reach than native posts because the ranking model predicts people will leave X, which looks like a shadowban but isn't one.

Sources

  1. Open-sourcing the For You timeline , X Open Source, August 13, 2026
  2. x-algorithm (For You feed source code, visibility filtering and Under the Hood labels) , xAI / X, GitHub
  3. underTheHoodLabels.strato (public list of visibility labels and their effects) , xAI / X, GitHub
  4. Freedom of Speech, Not Reach: An update on our enforcement philosophy , X Blog, April 17, 2023
  5. Our range of enforcement options , X Help Center
  6. Inside X's newly published visibility filtering code , tech-ish, August 14, 2026
Last updated
Writer, Twigor Blog

Web developer with nine years of WordPress work. Writes the Twigor blog's guides on growing and protecting accounts on X.