twittertroll

Guide

How to Spot Bot Networks on X

One fake account is hard to prove. Twenty that behave the same way are easy to see once you know where to look.

Short answer

Bot networks on X give themselves away through patterns across accounts: handles built from a name plus a long number, accounts created in the same weeks, identical or near-identical replies, posts landing within seconds of each other, and profile photos that are stock, stolen or generated. Search for a repeated phrase in quotes to see the whole group at once.

Key takeaways

  • Look at the group, not one account. Coordination is the tell.
  • Searching a suspicious reply's exact wording in quotes often exposes dozens of copies.
  • Same join month plus same posting rhythm across accounts is a strong signal.
  • Report the accounts as spam and block; don't reply.

What does a bot network look like?

A group of accounts that look different on the surface but behave almost identically.

The individual accounts may have different names and photos. Lay them side by side and the similarities jump out: created in the same few weeks, handles following the same pattern, bios built from the same template, posting at the same moments, pushing the same link or slogan.

Which signs should I check?

Check the handles, the dates, the wording and the timing across several accounts.

Sign What to look for
Handles Name plus 8 or more digits, or the same word with different numbers
Join dates Several accounts that joined in the same month
Photos Stock images, celebrities, or generated faces with odd ears, teeth or backgrounds
Bios Same structure or emoji pattern, or empty
Wording Identical replies, or the same sentence with one word swapped
Timing Replies landing within seconds of each other, around the clock
Ratios Following thousands, few followers, little engagement on own posts
Content Same link, coin, giveaway or "DM me" offer

The troll and bot checklist scores one account at a time; run it on three or four from the same swarm and compare.

How can search expose a network?

Search the repeated phrase in quotes and the copies line up in front of you.

Copy a suspicious reply's most distinctive sentence and search it with quotes on the Latest tab: "this changed my life check my pinned". If thirty accounts posted the same words, you've found the network. Other useful searches:

  • url:suspicious-site.com lists every post that links to the same domain.
  • to:bigaccount "exact phrase" shows the same reply pasted under one account's posts.
  • "phrase" since:2025-10-01 until:2025-10-02 shows how fast the copies arrived.

The search builder builds these from a form.

What should I do when a swarm hits my replies?

Don't engage, report a few accounts as spam, block them and limit replies on the post.

Replying to bots boosts them. Report the most obvious accounts as spam and block them. If the swarm keeps coming, change who can reply on the post and turn on notification filters for new accounts and accounts without a profile photo. Tell your followers not to click links in the replies.

Can the numbers alone prove anything?

No. Numbers point you at accounts worth a closer look.

An account can follow thousands of people for innocent reasons, and some real people post all day. Use the stats calculator for ratios and posting rate, then read the replies. Coordination across many accounts is the evidence that counts.

Questions people ask

Are all automated accounts against X's rules?

No. X allows automated accounts that follow its automation rules, such as weather or news feeds, and it offers a label for them. Coordinated fake accounts, fake engagement and spam are what the rules prohibit.

Can a bot network be made of real people?

Yes. Coordinated campaigns sometimes use real accounts working from the same script, which X's policies also cover as platform manipulation. They show many of the same patterns.

Why do bots reply to big accounts?

Replies under popular posts get seen by large audiences. Spam and scam networks reply quickly under viral posts to catch that traffic.

Does a default handle mean an account is a bot?

No. Lots of real people keep the username X suggested. It becomes meaningful when many accounts in the same swarm share that pattern along with other signs.

Sources

  1. Authenticity policy (X Help Center)
  2. Automation rules (X Help Center)