# Coordination Groups Sometimes abuse isn't a collection of individuals; it's a campaign. [Coordination Groups](https://clarify.net.au/accounts/groups) is where Clarify identifies clusters of accounts acting together: bot farms, brigading rings, and organised harassment.  ## How groups are found Two things work together. On every scan, Clarify's own checks look for hard evidence across recent comments: the same text from different accounts, several flagged accounts hitting one post within minutes, a burst of accounts Clarify had never seen before all arriving on the same post, the same obscure link pushed by a few accounts, and the same username on two platforms. Each finding is a **receipt**. Then the AI weighs those receipts, together with what the comments say, and proposes groups. Suggestions arrive on their own after a scan that turns up enough new evidence, or when you click **Run AI Analysis**. Either way they are only suggestions: nothing changes until you apply one. ## Reviewing a suggestion Each card shows who suggested it (**AI suggested**, or **From Clarify's checks** when the evidence alone was strong enough), the proposed members, a strength pill, and the receipts. Open a receipt to see the actual post and the comments behind it, with the gap between each comment and a link to open any of them in the comment queue. That is the evidence; judge it, not the summary. - **Apply** creates the group, or adds the accounts to an existing group when the card says it will. - **Edit** lets you rename the group, pick a type, and untick anyone who doesn't belong before applying. - **Dismiss** rejects the suggestion. A card marked **Narrative only** means Clarify's checks found nothing concrete linking the accounts; the AI thinks they share a script. Treat those with more caution. ## Inside a group Open a group to see: - **What connects them**, the receipts between the members, each opening to the comments behind it. - **Members**, each with a suspicion score and their comment history on your Page. - **Suspected coordinator**, when the analysis identifies an account that appears to be directing the others, with its reasoning. - **Combined activity**, the group's comments interleaved on one timeline. Comments posted within 30 minutes of each other are badged, which makes coordinated waves easy to see. Tick **Only comments in the receipts** to narrow it to the evidence. ## The group report Click **Quick Report** to generate a formal, printable dossier: summary, what connects the members (every receipt with its comments), member profiles with scores and signals, and the combined comment timeline. Use **Print / Save PDF** to produce a file suitable for handing to a platform, the eSafety Commissioner, or police. The report is timestamped and marked confidential. ## Escalating and tidying up - **Raise case** opens a case for the whole group on the [case board](https://clarify.net.au/cases) (harassment groups default to High priority). - **Dissolve group** removes the grouping if it turns out to be a false positive. The accounts stay tracked individually in the [Account Tracker](https://clarify.net.au/accounts).