What does SPAM_HIGH_RECALL mean on X?

X’s automated spam check flagged the post as one that may contain spam. The effect: the post is hidden from recommendations to people who don’t follow you. Your followers still see it. “High recall” means the check is built to catch a lot, so it also catches some normal posts.

What is the difference between SPAM_HIGH_RECALL and SpamHighRecall?

SPAM_HIGH_RECALL is a post label: one post is held back. SpamHighRecall is an account label: X thinks the whole account is likely to post spam, so its posts drop out of For You, even for followers. They still see you in Following. The account label is the one that stalls growth.

What does DoNotAmplify mean on X?

As an account label, it means X’s systems think the account may break its Terms of Service, so its posts are hidden from recommendations to non-followers while X reviews it. As a post label (DO_NOT_AMPLIFY), it is the older name for a malicious-link flag that X is replacing with MALICIOUS_URL.

Where do I see which labels X put on my account?

In X’s Under the Hood report at x.com/i/jf/under_the_hood, while logged in. There is no menu link to it. You need an account at least one year old and 10 or more posts in the prior month.

Is this the full list of X labels?

It is every label the Under the Hood report can show, copied from the allowlist in X’s open-source code. X uses more signals internally than the report shows, so a clean report is good news, not a full all-clear.

Many do, according to X’s open-source code. The adult-content account labels last 7 days by default. Most spam account labels from X’s models last 30 days. A pinned post with a bad link gives you SpamHighRecall for 1 week. The report also counts labels that were later removed, so one in your report may already be gone.

Can I appeal a label?

Not from the report or this page. For a post with a visible “limited visibility” notice, X’s appeal form for limited posts is the route. For most automated labels the fix is to change what likely caused it and check next month’s report.

All free tools

X Label Lookup: What Your Under the Hood Labels Mean

Type a label from your report, like SPAM_HIGH_RECALL. See X’s own wording, what it hides in plain English and what to do about it.

Label from your report

Try:

Nothing leaves your browser. Every label comes from X’s open-source code.

Where these labels come from

X’s Under the Hood report lists the labels that limited your posts and account last month. It prints a short description for each, and that copy lives in X’s open-source code. This page shows that same text word for word, then explains it plainly and says what to do. How to open and read your report: X Under the Hood report explained.

Label list checked against X’s code on October 17, 2026.

How bad is my label?

Hidden from non-followersYour followers still see it. X just stops showing it to new people.
Behind a warningShown behind a content warning, and kept from non-followers and minors.
Profile onlyOnly visible on your profile, with a public “limited visibility” notice.
Not shown on XNot shown on X at all, at least until a review or until you delete it.
Blocked in some countriesBlocked in the listed countries only. Elsewhere nothing changes.

An account label always outweighs a post label. One labeled post is one weak post. An account label can hold back everything you publish.

How the report counts labels

The code that builds the report is open too. Seven things it shows that the report doesn’t tell you:

Only your own original posts count. Reposts don’t, and an edited post counts once.
A post label counts if it was applied in the post’s first 7 days. Labels added later are not counted.
The count includes labels X removed again later, so a label in your report may already be gone.
An account label counts on a day only if it was active at midnight UTC. One added and removed the same day never shows.
There is no minimum. One labeled post is enough to show up.
A month’s report appears about 10 days after the month ends. Until then you see the month before.
If the post you reply to or quote gets dropped from For You, your post is dropped there too.

Post labels

Labels on a single post. They change how far that one post travels.

`SPAM_HIGH_RECALL`

Post labelHidden from non-followers

An automated spam check flagged this post as possibly spam. Your followers still see it, but X won’t recommend it to anyone else.

What X’s report says

What it means

Post detected by automated systems as one that may contain spam.

Effect on visibility

Post hidden from recommendations to non-followers.

What X’s code adds

Dropped from For You for people who don’t follow you, and from the video feed. Followers still see it in Following (visibility-filtering/rules/tweet_rules.rs).
Common triggers in the code: engagement bait and engagement farming found by Grok’s review, gibberish posts, “fast reply spam”, and links that pass through a low-quality site.
The fast-reply-spam and gibberish versions expire after 30 days (abuse-enforcement-service rules).

Read from X’s open-source code. X’s live settings can differ.

What to do

“High recall” means the check is built to catch a lot, so it also catches some normal posts. One or two in a month is noise. If you see many, look for what they share: the same reply pasted under many posts, lots of @mentions or hashtags, or a link in every post.

Full explainer for SPAM_HIGH_RECALL

`SPAM`

Post labelNot shown on X

X decided this post is spam under its Authenticity policy, so it is not shown at all.

What X’s report says

What it means

Post identified by automated systems or in response to a user report as likely to contain content in violation of X’s Authenticity policy.

Effect on visibility

Post is not shown on X.

What X’s code adds

Dropped for everyone but you, in Following too, not only in recommendations (visibility-filtering/rules/tweet_rules.rs).
One path in the code: a link in the post that X’s link check rates as BAD. Removing that verdict removes the label.

Read from X’s open-source code. X’s live settings can differ.

What to do

Stop whatever produced it: copy-paste replies, mass mentions, engagement swaps or automation that posts for you. Delete similar posts. Repeated spam is how a post label turns into an account label.

Full explainer for SPAM

`MALICIOUS_URL`

Post labelHidden from non-followers

The post links to a site X thinks may be harmful, so only your followers see it in their feeds.

What X’s report says

What it means

Post detected by automated systems as likely to contain a link to a malicious site.

Effect on visibility

Post hidden from recommendations to non-followers.

What X’s code adds

Applied together with DO_NOT_AMPLIFY when a link resolves to a site rated unsafe, and removed again if that rating is lifted.
A repost of your post gets the label on the original.
Dropped from For You for non-followers and from the video feed.

Read from X’s open-source code. X’s live settings can differ.

What to do

Check the link and any redirect or shortener in front of it. A hacked site or a spammy shortener can trigger this. Link straight to the real page, and remove links you don’t trust.

Full explainer for MALICIOUS_URL

`DO_NOT_AMPLIFY`

Post labelHidden from non-followers

The older name for MALICIOUS_URL. Same effect: the post is not recommended to non-followers.

What X’s report says

What it means

Historically applied to posts detected by automated systems as likely to contain a link to a malicious site. Being replaced by MALICIOUS_URL label and then will be fully removed.

Effect on visibility

Included in this report since any posts that have this label would be hidden from recommendations to non-followers.

What X’s code adds

Applied in the same step as MALICIOUS_URL, for links to sites rated unsafe, and removed with it.
Dropped from For You for non-followers and from the video feed.

Read from X’s open-source code. X’s live settings can differ.

What to do

Treat it like MALICIOUS_URL: check the link in the post and drop it if you are not sure it is safe.

Full explainer for DO_NOT_AMPLIFY

`NSFW_HIGH_RECALL`

Post labelHidden from non-followers

An automated check thinks the post may be adult content. Non-followers, minors, people with no age set and logged-out visitors won’t see it.

What X’s report says

What it means

Post detected by automated systems as one that may contain content subject to X’s Adult Content policy.

Effect on visibility

Post hidden from recommendations to non-followers and from underage users, users without a stated age, and logged out users.

What X’s code adds

Dropped from For You for non-followers. For logged-out visitors, minors and people with no age set, it only drops posts that have media.
The code shows Grok’s policy review applying it to sexual content, with or without media.
No expiry found in the code.

Read from X’s open-source code. X’s live settings can differ.

What to do

If the post is not adult content, check what could read that way to an automated system (an image, a crop, a word) and avoid it. If you do post adult content, mark your media as sensitive in Settings so X labels it for you.

Full explainer for NSFW_HIGH_RECALL

`NSFW_HIGH_PRECISION`

Post labelBehind a warning

X is fairly sure the post is adult content. It sits behind a warning, and non-followers, minors and logged-out visitors don’t see it.

What X’s report says

What it means

Post detected by automated systems or in response to a user report as likely to contain content subject to X’s Adult Content policy.

Effect on visibility

Post shown behind content warning, hidden from recommendations to non-followers and hidden from underage users, users without a stated age, and logged out users. In India, this content is shown as restricted due to local law(s).

What X’s code adds

Blurred in your followers’ feeds unless they allow sensitive media, and dropped from For You for non-followers.
If 3 of your last 5 posts get it, your whole account gets the NsfwHighPrecision label for 7 days. Reposts count too (safety-label-user-agg rules).
In the EU, UK, Australia and Brazil, viewers who haven’t verified their age get an age check first.

Read from X’s open-source code. X’s live settings can differ.

What to do

Full explainer for NSFW_HIGH_PRECISION

`NSFW_TEXT`

Post labelHidden from non-followers

The words in the post read as sexually explicit to X’s systems. Only followers who are verified adults see it.

What X’s report says

What it means

Post detected by automated systems as likely to contain adult sexually explicit language.

Effect on visibility

Post hidden from recommendations to non-followers, and hidden from underage users, users without a stated age, and logged out users.

What X’s code adds

Dropped for logged-out visitors, minors and people with no age set, even without media.
Unclear: the public For You rules have no line for this label, although the report says it hides the post there. That part may live in code X hasn’t published.

Read from X’s open-source code. X’s live settings can differ.

What to do

Look at the wording, including jokes and slang. If the post doesn’t need it, rephrase.

Full explainer for NSFW_TEXT

`NSFW_CARD_IMAGE`

Post labelBehind a warning

The preview image of a link in the post looks like adult content. The post sits behind a warning and is hidden from non-followers.

What X’s report says

What it means

Post detected by automated systems as likely to contain content subject to X’s Adult Content policy.

Effect on visibility

Post shown behind content warning and hidden from recommendations to non-followers and hidden from underage users, users without a stated age and logged out users.

What X’s code adds

The label sits on the link, not on you: when a link’s preview image is rated adult, every post sharing that link gets it (up to 5,000 posts per event).
The rating on the link lasts 7 days in the code. When it is removed, the label comes off the posts.

Read from X’s open-source code. X’s live settings can differ.

What to do

Check the preview image the linked page sets (its og:image). If it is your page, change it. If not, drop the link or post without the preview.

Full explainer for NSFW_CARD_IMAGE

`GORE_AND_VIOLENCE_HIGH_PRECISION`

Post labelBehind a warning

X thinks the post shows graphic violence. It sits behind a warning and only reaches adult followers.

What X’s report says

What it means

Post detected by automated systems or in response to a user report as likely to contain content subject to X’s Violent Content policy.

Effect on visibility

Post shown behind content warning, hidden from recommendations to non-followers and hidden from underage users, users without a stated age, and logged out users.

What X’s code adds

Applied by an image and video model (image score above 0.6, video at or above 0.95) or by Grok’s review of media posts.
Some large and gray-verified accounts go to a human review instead of an automatic label.
Blurred as “Violence” for viewers who don’t allow sensitive media, and dropped from For You for non-followers.

Read from X’s open-source code. X’s live settings can differ.

What to do

If the post needs graphic media (news, for example), mark it as sensitive yourself. If it doesn’t, post without it.

Full explainer for GORE_AND_VIOLENCE_HIGH_PRECISION

`NSFW_ADMIN`

Post labelBehind a warning

This comes from your account, not the post: your account carries the NsfwAdmin label, so every post gets this treatment.

What X’s report says

What it means

Post written by an account with an NsfwAdmin user label, where the account was identified in response to a user report as primarily posting content subject to X’s Adult Content or Violent Content policies.

Effect on visibility

Post shown behind content warning, hidden from recommendations to non-followers and from underage users, users without a stated age, and logged out users.

What X’s code adds

Every post you write while your account has the NsfwAdmin flag is counted here, so the count follows your posting, not the content of each post.
Dropped from For You for non-followers. Media posts are also blurred in Following for viewers who don’t allow sensitive media.

Read from X’s open-source code. X’s live settings can differ.

What to do

See NsfwAdmin under account labels. Changing one post won’t remove this.

Full explainer for NSFW_ADMIN

Also shown as: NSFW_ADMIN_STAMPED

`PDNA`

Post labelNot shown on X

The post is hidden while X reviews it for a possible Terms of Service violation.

What X’s report says

What it means

Post detected by automated systems as likely to contain content in violation of X’s Terms of Service policy.

Effect on visibility

Post is not shown on X while pending further review.

What X’s code adds

Dropped for everyone but you, in Following too, while it is reviewed.

Read from X’s open-source code. X’s live settings can differ.

What to do

Wait for the review. If the post clearly breaks a rule, delete it yourself. Don’t repost the same content in the meantime.

Full explainer for PDNA

`BOUNCE`

Post labelNot shown on X

X found the post breaks its rules and wants you to delete it. Until you do, it stays hidden.

What X’s report says

What it means

Post detected by automated systems or in response to a user report to contain content in violation of X’s Terms of Service policy.

Effect on visibility

Post is not shown on X while pending deletion by the author.

What X’s code adds

Dropped for everyone but you, in Following too, until you delete it.
Uncertain: Grok’s review “bounces” posts that encourage self-harm, but the code doesn’t show if that is this label.

Read from X’s open-source code. X’s live settings can differ.

What to do

Delete the post (X usually asks for this at login), or appeal if you think X got it wrong. An unresolved one can lock your account to read-only (see ReadOnly).

Full explainer for BOUNCE

`FOSNR_ABUSE`

Post labelProfile only

Someone reported the post as harassment and X agreed. It only shows on your profile, with a visible “limited visibility” notice.

What X’s report says

What it means

Post detected in response to a user report to contain content in violation of X’s Abuse and Harassment policy.

Effect on visibility

Post discoverability is restricted to the author’s profile. All users can see a label explaining that the post has limited visibility.

What X’s code adds

Hidden from everyone but you in both Following and For You, not only from non-followers (visibility-filtering/rules/tweet_rules.rs).
You see a notice you can appeal, and 13 actions are switched off on the post: like, reply, repost, quote, share, bookmark, pin, copy link, DM, edit, highlight, embed and lists.
A repost of the post carries the same notice.

Read from X’s open-source code. X’s live settings can differ.

What to do

Reread the post against the X policy it names. If it breaks the rule, delete it. If you think the label is wrong, X’s appeal form for limited posts is the place to say so.

Full explainer for FOSNR_ABUSE

`FOSNR_ABUSE_INSULTS`

Post labelHidden from non-followers

The post reads as an insult aimed at someone. Followers still see it, with a “limited visibility” notice, but non-followers don’t get it recommended.

What X’s report says

What it means

Post detected by automated systems or in response to a user report to contain content in violation of X’s Abuse and Harassment policy.

Effect on visibility

Post hidden from recommendations to non-followers. All users can see a label explaining that the post has limited visibility.

What X’s code adds

Followers still see it in Following. Non-followers don’t get it in For You, and in some views followers see it behind a soft “abuse” notice.
Grok’s review applies it to “insults with a target”. In the code that check only runs on English posts.
Only editing is switched off for you, unlike the stricter FOSNR labels.

Read from X’s open-source code. X’s live settings can differ.

What to do

Heated replies are the usual source. Disagree with the idea, not the person.

Full explainer for FOSNR_ABUSE_INSULTS

`FOSNR_HATEFUL_CONDUCT`

Post labelProfile only

X found the post breaks its Hateful Conduct policy. It only shows on your profile, with a visible notice.

What X’s report says

What it means

Post detected by automation systems or in response to a user report to contain content in violation of X’s Hateful Conduct policy.

Effect on visibility

Post discoverability is restricted to the author’s profile. All users can see a label explaining that the post has limited visibility.

What X’s code adds

Hidden from everyone but you in both Following and For You, not only from non-followers (visibility-filtering/rules/tweet_rules.rs).
You see a notice you can appeal, and 13 actions are switched off on the post: like, reply, repost, quote, share, bookmark, pin, copy link, DM, edit, highlight, embed and lists.
A repost of the post carries the same notice.

Read from X’s open-source code. X’s live settings can differ.

What to do

Reread the post against the X policy it names. If it breaks the rule, delete it. If you think the label is wrong, X’s appeal form for limited posts is the place to say so.

Full explainer for FOSNR_HATEFUL_CONDUCT

`FOSNR_VIOLENT_SPEECH`

Post labelProfile only

A report led X to decide the post contains violent speech. It only shows on your profile, with a visible notice.

What X’s report says

What it means

Post detected in response to a user report to contain content in violation of X’s Violent Content policy.

Effect on visibility

Post discoverability is restricted to the author’s profile. All users can see a label explaining that the post has limited visibility.

What X’s code adds

Hidden from everyone but you in both Following and For You, not only from non-followers (visibility-filtering/rules/tweet_rules.rs).
You see a notice you can appeal, and 13 actions are switched off on the post: like, reply, repost, quote, share, bookmark, pin, copy link, DM, edit, highlight, embed and lists.
A repost of the post carries the same notice.

Read from X’s open-source code. X’s live settings can differ.

What to do

Reread the post against the X policy it names. If it breaks the rule, delete it. If you think the label is wrong, X’s appeal form for limited posts is the place to say so. Figures of speech (“I’ll kill this launch”) can get reported too.

Full explainer for FOSNR_VIOLENT_SPEECH

`FOSNR_CIVIC_INTEGRITY`

Post labelProfile only

A report led X to decide the post breaks its rules on elections and civic processes. It only shows on your profile, with a notice.

What X’s report says

What it means

Post detected in response to a user report to contain content in violation of X’s Civic Integrity policy.

Effect on visibility

Post discoverability is restricted to the author’s profile. All users can see a label explaining that the post has limited visibility.

What X’s code adds

Hidden from everyone but you in both Following and For You, not only from non-followers (visibility-filtering/rules/tweet_rules.rs).
You see a notice you can appeal, and 13 actions are switched off on the post: like, reply, repost, quote, share, bookmark, pin, copy link, DM, edit, highlight, embed and lists.
A repost of the post carries the same notice.

Read from X’s open-source code. X’s live settings can differ.

What to do

Reread the post against the X policy it names. If it breaks the rule, delete it. If you think the label is wrong, X’s appeal form for limited posts is the place to say so.

Full explainer for FOSNR_CIVIC_INTEGRITY

`BRAZIL_ELECTION_LEGAL`

Post labelBlocked in some countries

AI-made media about Brazil’s election, posted during the legal silence period. Only affects recommendations in Brazil.

What X’s report says

What it means

Post detected by automated systems as one that may contain AI generated media related to electoral processes in violation of Brazil local law during silence period, specifically Tribunal Superior Eleitoral Resolution n. 23.610, Art9-B, § 3-A

Effect on visibility

In Brazil, post hidden from recommendations.

What X’s code adds

Only applied when the viewer is in Brazil: dropped from For You and the video feed there.

Read from X’s open-source code. X’s live settings can differ.

What to do

Nothing changes outside Brazil. If Brazil is your audience, avoid AI-generated election media during the silence period.

Full explainer for BRAZIL_ELECTION_LEGAL

`FOR_EMERGENCY_USE_ONLY`

Post labelBehind a warning

A label X keeps for emergencies, like a fast-moving crisis. The post sits behind a notice and is kept out of Home.

What X’s report says

What it means

Emergency incident-response label with a custom notice.

Effect on visibility

Posts that have this label would be shown behind a notice and not shown in the Home timeline.

What X’s code adds

Dropped from the home feeds for everyone, you included.

Read from X’s open-source code. X’s live settings can differ.

What to do

Read the notice on the post. It should say what the emergency is about.

Full explainer for FOR_EMERGENCY_USE_ONLY

Account labels

Labels on your whole account. These matter most, because they affect every post.

`SpamHighRecall`

Account labelHidden from non-followers

Your whole account looks spammy to X’s systems, so your posts drop out of For You, even for your followers. They still see you in Following. This is the one that stalls growth.

What X’s report says

What it means

Account detected by automated systems as likely to post spam.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Your posts are dropped from For You, followers included, and from the video feed. Followers still see them in Following (author_rules.rs).
Most versions expire after 30 days, sometimes with a captcha first. They come from spam models, including one named “llm_slop_user” for AI-slop accounts.
Pinning a post with a link to a low-quality or bad site sets it for 1 week. Following people while that pin is up does too.

Read from X’s open-source code. X’s live settings can differ.

What to do

Change behavior before content: stop bulk or copy-paste replies, mass follows and unfollows, engagement pods and automation that acts for you. Unpin any post with a sketchy link. Most versions expire after 30 days, so post normally and check next month’s report.

Full explainer for SpamHighRecall

`DoNotAmplify`

Account labelHidden from non-followers

X’s systems think the account may break its rules, so its posts are not recommended until a review finishes.

What X’s report says

What it means

Account detected by automated systems as likely to contain content in violation of X’s Terms of Service policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers pending further review.

What X’s code adds

Your posts are dropped from For You for non-followers only, and from the video feed. Followers still get them.

Read from X’s open-source code. X’s live settings can differ.

What to do

Remove posts that could break the rules and pause anything automated. Then wait for the review.

Full explainer for DoNotAmplify

`AbusiveHighRecall`

Account labelHidden from non-followers

The account behaves like a bot to X’s systems. Posts aren’t recommended, and X asks you to pass a security check.

What X’s report says

What it means

Account detected by automated systems may be in violation of X’s Authenticity policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers. Security challenge required to prevent automated abuse.

What X’s code adds

Your posts are dropped from For You for non-followers only, and from the video feed. Followers still get them.

Read from X’s open-source code. X’s live settings can differ.

What to do

Complete the security challenge at login. Turn off tools that like, follow, reply or DM for you, and slow down any burst of actions.

Full explainer for AbusiveHighRecall

`ImpersonationHighPrecision`

Account labelHidden from non-followers

X thinks the account may be pretending to be someone else.

What X’s report says

What it means

Account detected by automated systems as likely in violation of X’s Authenticity policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Your posts are dropped from For You for everyone but you, followers included, and from the video feed. Following is not affected.

Read from X’s open-source code. X’s live settings can differ.

What to do

Make it clear who you are: your own name and photo, and “parody” or “fan” in the name and bio if that is what the account is.

Full explainer for ImpersonationHighPrecision

`Compromised`

Account labelHidden from non-followers

X thinks someone else got into the account.

What X’s report says

What it means

Account identified by automated systems or in response to a user report as compromised.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers. Password reset required.

What X’s code adds

Your posts are dropped from For You for everyone but you, followers included, and from the video feed. Following is not affected.

Read from X’s open-source code. X’s live settings can differ.

What to do

Reset your password, turn on two-factor login, remove connected apps you don’t recognize and delete anything you didn’t post.

Full explainer for Compromised

`ReadOnly`

Account labelHidden from non-followers

You can’t post until you delete a post X says breaks its rules.

What X’s report says

What it means

Account locked to read-only due to violation of X’s Terms of Service policy that requires post removal from the platform.

Effect on visibility

The account can only read content on the platform until the post in violation is removed. The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Your posts are dropped from For You for everyone but you, followers included. Following is not affected.

Read from X’s open-source code. X’s live settings can differ.

What to do

Full explainer for ReadOnly

`NsfwAdmin`

Account labelBehind a warning

After a report, X decided the account mainly posts adult or violent content. Every post now sits behind a warning.

What X’s report says

What it means

Account identified in response to a user report as primarily posting content subject to X’s Adult Content or Violent Content policies.

Effect on visibility

The account’s posts are shown behind content warning, hidden from recommendations to non-followers and from underage users, users without a stated age, and logged out users.

What X’s code adds

All your posts are dropped from For You for non-followers, text-only posts included.
Media posts are blurred in Following for viewers who don’t allow sensitive media, and hidden from logged-out visitors and minors.

Read from X’s open-source code. X’s live settings can differ.

What to do

If that is wrong, appeal. If the account does post that content, this is X working as intended and growth to a general audience will stay limited.

Full explainer for NsfwAdmin

`NsfwHighPrecision`

Account labelHidden from non-followers

X is fairly sure the account posts adult content, so its posts aren’t recommended.

What X’s report says

What it means

Account detected by automated systems as likely to post content subject to X’s Adult Content policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Set by a media classifier when you post more than 2 adult-media posts within about a day. “Near perfect” means a score above 0.99 (ApplyNsfwUserLabel.df).
It expires: 7 days by default, 3 days for some large accounts (ApplyNsfwUserLabelOrCreateReport.df).
Followers still see your posts in Following. The drop is in For You, and it applies to followers there too.
Also set for 7 days when 3 of your last 5 posts carry NSFW_HIGH_PRECISION.

Read from X’s open-source code. X’s live settings can differ.

What to do

Full explainer for NsfwHighPrecision

`NsfwHighRecall`

Account labelHidden from non-followers

A broad automated check thinks the account may post adult content.

What X’s report says

What it means

Account detected by automated systems as one that may post content subject to X’s Adult Content policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Set by a media classifier when you post more than 2 adult-media posts within about a day. “Near perfect” means a score above 0.99 (ApplyNsfwUserLabel.df).
It expires: 7 days by default, 3 days for some large accounts (ApplyNsfwUserLabelOrCreateReport.df).
Followers still see your posts in Following. The drop is in For You, and it applies to followers there too.

Read from X’s open-source code. X’s live settings can differ.

What to do

Full explainer for NsfwHighRecall

`NsfwNearPerfect`

Account labelHidden from non-followers

Another automated adult-content check on the account as a whole.

What X’s report says

What it means

Account detected by automated systems as one that may post content subject to X’s Adult Content policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Set by a media classifier when you post more than 2 adult-media posts within about a day. “Near perfect” means a score above 0.99 (ApplyNsfwUserLabel.df).
It expires: 7 days by default, 3 days for some large accounts (ApplyNsfwUserLabelOrCreateReport.df).
Followers still see your posts in Following. The drop is in For You, and it applies to followers there too.

Read from X’s open-source code. X’s live settings can differ.

What to do

Full explainer for NsfwNearPerfect

`NsfwAvatarImage`

Account labelHidden from non-followers

Your profile picture looks like adult content to X’s systems, and that limits every post.

What X’s report says

What it means

Account detected by automated systems as one that may post content subject to X’s Adult Content policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Your posts are dropped from For You, followers included, and from the video feed. Following is not affected.
The code doesn’t show what sets it. A violent avatar goes to a human review instead.

Read from X’s open-source code. X’s live settings can differ.

What to do

Change your profile picture. It is one of the cheapest fixes on this list.

Full explainer for NsfwAvatarImage

`NsfwBannerImage`

Account labelHidden from non-followers

Your header image looks like adult content to X’s systems, and that limits every post.

What X’s report says

What it means

Account detected by automated systems as one that may post content subject to X’s Adult Content policy.

Effect on visibility

The account’s posts are hidden from recommendations to non-followers.

What X’s code adds

Your posts are dropped from For You, followers included, and from the video feed. Following is not affected.
The code doesn’t show what sets it. A violent header image goes to a human review instead.

Historically applied to accounts withheld in response to a legal demand or if found to violate local law(s) following a third-party report.

Effect on visibility

This account’s posts are not shown on X in {countries} where it violates local law(s).

What to do

Check the country codes in brackets.

Full explainer for UnspecifiedReason

What this lookup cannot tell you

Which of your posts got a label. The report gives counts, not post links.
Signals X keeps internal. The report only covers this list, so a label not here never shows up in it.
Why a specific post was flagged. Automated checks don’t explain themselves.

Not old enough for a report yet? The Twitter shadowban checker tests search and reply visibility from the outside.

X labels: common questions

What does SPAM_HIGH_RECALL mean on X?: X’s automated spam check flagged the post as one that may contain spam. The effect: the post is hidden from recommendations to people who don’t follow you. Your followers still see it. “High recall” means the check is built to catch a lot, so it also catches some normal posts.
What is the difference between SPAM_HIGH_RECALL and SpamHighRecall?: SPAM_HIGH_RECALL is a post label: one post is held back. SpamHighRecall is an account label: X thinks the whole account is likely to post spam, so its posts drop out of For You, even for followers. They still see you in Following. The account label is the one that stalls growth.
What does DoNotAmplify mean on X?: As an account label, it means X’s systems think the account may break its Terms of Service, so its posts are hidden from recommendations to non-followers while X reviews it. As a post label (DO_NOT_AMPLIFY), it is the older name for a malicious-link flag that X is replacing with MALICIOUS_URL.
Where do I see which labels X put on my account?: In X’s Under the Hood report at x.com/i/jf/under_the_hood, while logged in. There is no menu link to it. You need an account at least one year old and 10 or more posts in the prior month.
Is this the full list of X labels?: It is every label the Under the Hood report can show, copied from the allowlist in X’s open-source code. X uses more signals internally than the report shows, so a clean report is good news, not a full all-clear.
Do X labels expire?: Many do, according to X’s open-source code. The adult-content account labels last 7 days by default. Most spam account labels from X’s models last 30 days. A pinned post with a bad link gives you SpamHighRecall for 1 week. The report also counts labels that were later removed, so one in your report may already be gone.
Can I appeal a label?: Not from the report or this page. For a post with a visible “limited visibility” notice, X’s appeal form for limited posts is the route. For most automated labels the fix is to change what likely caused it and check next month’s report.