Suspended on X (Twitter) for inauthentic behavior: what X says and what to do

What X means by inauthentic behavior

X’s authenticity policy has three parts: inauthentic accounts, inauthentic behaviors and inauthentic content. Suspensions for “inauthentic behavior” come from the middle one, which opens like this:

You may not engage in behaviors that manipulate X or artificially impact how content is discovered and amplified.

Under it sit two lists that matter to anyone who replies a lot.

Content spam

You may not share or post content in a bulk, duplicative, irrelevant or unsolicited manner that disrupts people’s experience.

On X’s list of examples: “sending bulk, aggressive, high-volume unsolicited replies, mentions, or direct messages”, replying to promote something that has nothing to do with the post, and posting the same thing again and again, which X calls copypasta.

Engagement spam

We prohibit inauthentic use of X engagement features to artificially impact traffic or disrupt people’s experience.

Here X lists swapping likes, replies, reposts or follows with other people, following and unfollowing in bulk to grow a follower count, “engaging with posts aggressively or through the use of automation”, and paying a service to do any of it for you.

The other parts of the policy can also end in a suspension: automated accounts outside X’s developer rules, fake personas, and running several accounts to boost each other. X adds that you are “ultimately responsible for third-party applications” you let into your account.

The reply habits that look like it

None of these needs a bot. We did most of them by hand, with one account that got suspended and another that got a label on its reach.

What X can do about it

X says “the consequences for violating this policy depend on the severity of the violation as well as any previous history of violations”, and lists what it can do:

• Anti-spam challenges. The account is locked and you’re asked for a phone number or to solve a puzzle.
• Restricting reach. The label we got: your posts and replies are left out of search, trends, notifications and other people’s threads.
• Temporary loss of access to features, such as posting or Direct Messages.
• Suspension.

For severe violations, accounts will be permanently suspended at first detection.

For a first offense, X says it may instead ask you to delete posts or lock the account for a while. What each of these states looks like from the inside is in X account locked, limited or suspended.

X gets it wrong too

In March 2026, many people on X reported sudden suspensions for inauthentic behavior. Nikita Bier, then X’s head of product, explained what happened:

Fixed. 99% of those suspensions were reverted today. For about 12 hours, a new spam filter had falsely tagged subset of accounts.

In the same post he wrote that “spam has become the greatest risk to the platform”, and that X is upgrading its tools as AI gets better. A suspension can be a mistake, and an appeal is how X finds out.

How to appeal an inauthentic behavior suspension

1. Log in to the suspended account. If X asks you to confirm a phone number or email, do it. Some suspensions lift that way.
2. If the account is still suspended, file an appeal from it. X’s form, Appeal a locked or suspended account, only works while you are logged in to the suspended account. X’s instructions are to log in first, then open the form in a new tab.
3. Say plainly what happened. Who runs the account, what you post, and anything that may have looked like spam, such as an app you have since removed or a burst of replies in one evening.
4. Then wait. X doesn’t say how long a review takes.

From our own account: the first one we lost was suspended after weeks under a label. We appealed several times, and every appeal was turned down. The full story is in We lost a Twitter (X) account because of replies.

Don’t open a new account to replace it

X’s policy names this directly:

If an account has been suspended for violations of the X Rules, X reserves the right to also suspend any other account we believe the same account holder or entity may be operating in violation of our earlier suspension, regardless of when the other account was created.

The same section lists “creating new accounts” and “imitating a suspended account to replace it” as ban evasion. An appeal is the way back.

How to keep replying without looking like spam

• Reply to fewer people, better. One reply about the post beats five that could sit under anything.
• Spread your replies out. A steady pace through the day reads as a person. Fifty in an hour reads as a bulk run.
• One post per person per day. Pick the post you have something to say about.
• Write each reply for its post. No templates, no pasted lines.
• Leave out apps that act for you, and remove any you no longer use from your account’s connected apps.

Twick came out of the mistakes above. It drafts replies about the post in front of you, in your voice, and you press Reply on X yourself. Safe mode reminds you when you pass 30 replies in an hour, 100 in a day, 5 to the same person in a day, or two replies only seconds apart. Twick can’t promise X will never suspend an account, and no tool can. It helps you avoid the habits on X’s list.

Questions

Can I get my account back after an inauthentic behavior suspension?

It happens. X restored 99% of the accounts its March 2026 spam filter flagged by mistake. When the suspension is right, an appeal rarely changes it: ours never did.

How long does an appeal take?

X doesn’t say.

Is “platform manipulation” the same thing?

Close to it. Our label notice said our account “may contain spam or be engaging in other types of platform manipulation”. X’s help pages on platform manipulation now lead to the same authenticity policy.

Does X Premium protect an account?

X doesn’t say it does. Premium gives replies a lift in the ranking, but our Premium account still got a label on its reach.

Can a reply tool get my account suspended?

X holds you responsible for the apps you authorize. An app that posts, likes or follows on its own is what the policy calls automation. Twick never does any of that: it puts a draft in the reply box, and you press Reply.