Telegram blocklists: keep member bans local and reviewable
Maintain a narrow blocked-member policy, distinguish account bans from link and word lists, review old decisions and restore access without broad exemptions.
A useful blocked-member policy answers a narrow question: which accounts are currently excluded from this group, for what reason, and who can review that decision? It does not need to become a permanent reputation database about everyone your community encounters.
The phrase “Telegram blacklist” is also used for prohibited websites and word lists. Putting them under one label leads to practical mistakes: a moderator removes a domain entry and expects a member's ban to disappear, or bans a person when the actual problem is a recurring link. Start by identifying what the list contains.
Four lists with different purposes
| List | What it contains | What an entry controls |
|---|---|---|
| Banned members of a group | Telegram accounts | Access to that group |
| Link blocklist | Domains or supported Telegram sources | How matching links are handled by the filter |
| Stop-word list | Words, phrases or supported matching rules | Which messages match that list's configured consequences |
A person's private blocked contacts | Accounts blocked by that person | Their personal contact boundary, not the group's admission policy |
A filter match can lead to a member restriction when its configured action or warning threshold calls for one. The relationship is a sequence of actions, not a shared list: removing the original filter entry does not undo a restriction that has already been applied.
Write the policy before collecting names
Decide which conduct justifies exclusion and which cases deserve a warning, a posting pause or review. Use observable behavior tied to a published rule. Similar profile photos, unfamiliar names, account silence or membership in another community are not enough to establish that someone committed the same violation.
For a hypothetical open-source support group, a narrow policy could be:
- unsolicited sales pitches receive a rule explanation before escalation, unless the content creates an immediate safety risk;
- repeated promotion after a clear instruction may result in a timed or indefinite local ban;
- a disputed decision goes to another authorized moderator where possible;
- an indefinite ban has no automatic end, but can still be reviewed when relevant new information appears.
A review date belongs to your moderation process. Writing “review in a month” in an incident note does not create a timed Telegram ban or schedule an automatic unban.
Keep a small, private decision record
Telegram's actual member state is the source to check for whether a ban is active. A private team record supplies the reasoning and accountability that a list of account names cannot.
For each decision, keep only what is useful:
- the affected group and the verified account identifier;
- the action, decision time and actual expiry, if one exists;
- a short rule-based reason and a reference to necessary evidence;
- who made the decision and who owns an appeal;
- the outcome of the latest review, including any reversal.
This is a recommended team practice, not a claim that Defendy provides a case-management form or a user-ban spreadsheet importer. Store records in an existing restricted moderator workspace. Avoid public accusation lists, unrelated personal details and indefinite copies of whole conversations. Limit retention to what the community genuinely needs for unresolved or recurring cases.
Add or remove a member restriction safely
Defendy's supported commands include /ban, /tban, /kick and /unban. Their purposes differ: kick permits a later return; a ban excludes the account until it is lifted or its applicable duration ends. The ban guide covers choosing the action and its message-history risks.
Reply to the verified member's own message, or use a supported explicit username or numeric ID. Do not combine a reply to one account with another account's identifier: the reply takes priority. For a timed ban without a reply, duration comes first, for example /tban 1d @member_username repeated promotion. This is illustrative syntax; replace the placeholder only after verifying the target.
Use /unban when a reviewed decision calls for restoring eligibility to join. Then confirm the actual state and provide the normal admission route. Unbanning does not itself bring the person back or clear warning points. If the decision originated in a faulty filter, use the recovery checklist before asking them to try again.
The reviewed command set does not establish a /banlist report with arbitrary filters, /massban, or a CSV editor for banned users. Do not build an operating procedure around commands copied from another bot's manual. Check banned or removed members through Telegram's member-management controls; their labels can vary by client.
Keep source lists separate from member policy
Link blocklists
In Defendy, open Anti-spam → Links for the selected group. Websites and Telegram channels/chats have separate modes. The exact English mode label Blacklist means that the ordinary source check removes only listed sources. It is not a list of people who are forbidden to participate.
Popular-site exceptions can be evaluated before a source blocklist. Invitation and group-quote checks have separate behavior. Therefore, a listed domain that remains allowed, or an unlisted invitation that is removed, needs a rule-order check rather than a member ban. Use the link-filter reference and test the specific format.
The exact punishment label None does not switch message deletion off. Inspect the filter's consequences before using a normal member's useful message as a test. Prefer disposable test content in an isolated group.
Stop-word lists
A named stop-word list describes message content, with its own deletion and punishment choices. A phrase may be prohibited in advertising but legitimate in a question or quotation. Test safe examples alongside intended matches before introducing member penalties.
The current Global spam word list control is unavailable in the editor. Its presence as a planned setting does not establish an active shared word list, and it is not a shared database of banned accounts. See stop-word behavior and limitations.
Review stale decisions without silently broadening access
Review a case when its agreed review date arrives, a credible appeal appears or the rule that caused it changes. Compare the original reason with current policy and current Telegram state. A timed ban may already have expired while your note still says “blocked.” Conversely, correcting a note does not change Telegram permissions.
If exclusion remains justified, record the decision without adding unrelated claims about the person. If it is reversed, remove the actual restriction, resolve any mistaken warning points, and confirm that normal joining and posting work. Do not grant administrator status or broad filter immunity merely to fix one false positive.
For several groups, make each community's scope and decision owner explicit. A ban in one Defendy group does not establish a ban in another. Do not assume an automatic federation, universal Telegram blacklist or shared appeal mechanism. If another community shares an allegation, treat it as information to assess within your policy, not an instruction to import punishments.
Sources and verification
Checked on 5 October 2026. References: the Defendy command reference, link settings and stop-word guide, and Telegram's chat-specific ban method. The review does not verify the version running in a particular group. The policy example is hypothetical.
