X sensitive media flags and how they limit reach

X lets people control whether they want to see media that has been marked sensitive. On the posting side, there is an account setting to mark the media you post as containing material that may be sensitive. When that setting is on, your images and videos go out with that expectation. You might have turned it on once and forgotten it. It is the first thing to check, because it is under your control and it does not require a support ticket.

Separately, X labels individual posts. The underlying post object that third-party clients and checkers look at has long included a possibly_sensitive field. When that is true, clients are supposed to warn or withhold the media depending on the viewer's setting. You might see an interstitial on the web. Someone else, who disabled sensitive media, may see a placeholder instead of the picture. The text of the post can still be there.

Automated systems apply these labels. They are not a human telling you that you are shadowbanned. They also make mistakes, especially around medical images, news photos, art, and anything that merely resembles nudity or violence. A wrong label is still a label: viewers who hide sensitive media will not get a normal impression of the post.

A label on one post does not relocate your handle out of search. If from:handle Latest is empty, you have a search problem to test on its own in the search ban guide. If replies are missing from threads, that is the ghost ban or reply deboost test. Sensitive media can sit beside those issues. It does not explain them by itself.

Open the post in a logged-out window, because your own account may be less aggressive about warning you. If the image is behind a click-through warning, the post is being treated as sensitive for that viewer. Ask a friend who has sensitive media turned off whether they see the image inline. If they see a block instead of the picture, their setting and your label are doing exactly what the product describes.

Reach drops for a boring reason on top of any ranking penalty: a lot of people do not click warnings. A post that needs an extra tap loses casual viewers. If several recent posts are flagged, your timeline starts to look empty to anyone who hides sensitive content. That can be misread as a shadowban when a logged-out search still returns the posts.

Some flagged posts are harder than a warning. People describe media that will not load, or posts that search will not return, while the profile still shows a tombstone or the text. Do not invent a single rule for every one of those states. Note what you can reproduce: warning, hidden media, or missing from search. Those are three different sentences, and only the third is a search-visibility claim.

View counters interact badly with this confusion. A warned post can show a low or delayed count because fewer people rendered it, or because the count itself is slow. A zero on a post that is two minutes old is not evidence of a flag. The way to separate those stories is in why posts get 0 views.

• You turned on the account setting that marks your media as sensitive. Every image inherits it until you turn the setting off.
• An automated classifier scored the picture or video as adult content, graphic violence, or something adjacent. News and medical posts get caught.
• A viewer report sent the post through a review path. A report is not a result you can read back. The label, if one appears, is the result.
• The post really does match X's sensitive media rules. In that case the warning is the product working, not a bug in a checker.

Text-only posts are a useful control. If your text posts are searchable and your images are the only ones behind warnings, you are debugging media labels, not an account-wide shadowban. If text posts are also missing from logged-out search, run that test properly instead of deleting every image.

Links do not set possibly_sensitive by themselves. A spammy link and a sensitive image are different inputs. Accounts often have both at once because the same campaign used sensational images and short links. Untangle them when you clean up, or you will 'fix' the wrong thing. Link and behavior cleanup lives in how to recover.

This site's checker includes a sensitive-media probe. It looks at recent posts and reports whether they carry a sensitive flag, and it can list samples when it finds them. That is a reading of a public field on those posts. It is not X confirming an enforcement case, and it is not a statement about your search ranking.

Read the card with the other cards. A sensitive warning plus a passing search card means the posts are flagged and still indexed. A sensitive warning plus a failing search card means you have two findings, and deleting one image may not repair search. An error on the sensitive card means those posts could not be read. Do not treat an error as 'everything is flagged.'

The probe only covers a slice of recent posts, not your entire archive. An old viral post can be flagged while the probe looks at this week's text posts and says little. If you care about a specific post, open that post logged out and look for the warning yourself. The manual test is still the way to check search and replies, which this flag does not answer.

No checker can see the private reason code behind a classifier. Under the Hood, for accounts that can open it, may show labels X applied over the prior month, without listing the post IDs. If you are eligible, that report is the first-party complement. It is explained in Under the Hood vs checkers.

1. In X settings, find the control that marks your media as sensitive. If you turned it on by accident, turn it off. New posts should stop inheriting the setting. Old posts may keep the label they were given.
2. Open the specific posts that show a warning. If the media does not need to be on X, delete those posts. If it is a mistaken label on a legitimate image, use X's own reporting or appeal path on that post rather than deleting the whole account.
3. Post one new text-only update and one new ordinary image after the setting is off. Later, check both logged out. You are looking for whether new media still warns. You are not looking for instant distribution.
4. Leave search and reply tests for the guides that match those symptoms. Do not assume a media warning will clear a from: miss.

Turning the setting off does not promise that X will remove an automated label from an existing post. Some labels stay until the post is deleted. That is a reason to be selective about deletion, not a reason to wipe the profile. Keep anything you still want searchable, and remove the posts whose media you are willing to lose.

There is no public countdown for a sensitive label. It is attached to the post, not to a weekend. How long shadowbans last talks about this difference: a behavior limit can ease while you wait, and a label on a live post can sit there until the post changes. Retest the post. Do not retest by uploading the same borderline image ten times.

If you tell someone you are shadowbanned when the only finding is a content warning, you will take the wrong next step. You might stop replying, buy a new account, or file a generic support ticket, while the setting that marks your media sensitive is still on. Name the flag.

Viewers differ. One person with sensitive media enabled will click through and say you are fine. Another with it disabled will say your posts are blank. Both can be telling the truth. Ask which setting they use before you average their screenshots into a ban.

Checkers that scan possibly_sensitive will also flag accounts that opted in on purpose, including artists and adult creators who want the interstitial. A positive probe is a description of the field, not an accusation. Read it next to the search and reply cards, then confirm on the post itself. When you want those cards in one pass, run the checker.