Wikipedia:Blocking policy

Blocking is the method by which administrators technically prevent users from editing Wikipedia. Blocks may be applied to user accounts, to IP addresses, and to IP address ranges, for either a definite or an indefinite time, to all or a subset of pages. Blocked users can continue to access Wikipedia, but cannot edit any page they are blocked from (including, if appropriate, their own user pages). In most cases, a site-wide blocked user will only be able to edit their own user talk page.

Blocks are used to prevent damage or disruption to Wikipedia, not to punish users (see ). Any user may report disruption and ask administrators to consider blocking a disruptive account or IP address (see ).

If editors believe a block has been improperly issued, they can request a review of that block at Administrative action review. Administrators can unblock a user when they feel the block is unwarranted or no longer appropriate.

Blocking is different from banning, which is a formal retraction of editing privileges on all or part of Wikipedia. Blocks disable a user's ability to edit pages; bans do not. However, bans may be enforced by blocks; users who are subject to a total ban, or who breach the terms of a partial ban, will most likely be site-wide blocked to enforce the ban.

Purpose and goals
Blocks serve to protect the project from harm, and reduce likely future problems. Blocks may escalate in duration if problems recur. They are meted out not as retribution but to protect the project and other users from disruption and inappropriate conduct, and to deter any future possible repetitions of inappropriate conduct. Blocking is one of the most powerful tools that are entrusted to administrators, who should be familiar with the circumstances prior to intervening and are required to be able to justify any block that they issue.

In general, once a matter has become "cold" and the risk of present disruption has clearly ended, reopening it by blocking retrospectively is usually not appropriate. In this situation, if an ongoing or serious concern persists, several dispute resolution processes exist to allow discussion and possible sanction of a user.

Blocks can be appealed (see ). Requests to be unblocked are also decided in light of prevention and deterrence. A user may be unblocked earlier if the user agrees to desist and appears to have learned from the matter, or if the situation was temporary and has now ended. Likewise, a user who has previously returned to inappropriate conduct after other unblocks may find their unblock request declined for deterrence reasons, to emphasize the importance of change and unacceptability of the conduct.

Blocks should not be punitive
Blocks should not be used:
 * 1) to retaliate;
 * 2) to disparage;
 * 3) to punish; or
 * 4) if there is no current conduct issue of concern.

Blocks should be preventative
Blocks should be used to:
 * 1) prevent imminent or continuing damage and disruption to Wikipedia;
 * 2) deter the continuation of present, disruptive behavior; and
 * 3) encourage a more productive, congenial editing style within community norms.

Deterrence is based upon the likelihood of repetition. For example, though it might have been justifiable to block an editor a short time ago, such a block may no longer be justifiable right now, particularly if the actions have since ceased or the conduct issues have been resolved.

Common rationales for blocks
The following are some of the most common rationales for blocks.

As a rule of thumb, when in doubt, do not block; instead, consult other administrators for advice. After placing a potentially controversial block, it is a good idea to make a note of the block at the administrators' incidents noticeboard for peer review.

Administrators should take special care when dealing with new users. Beginning editors are often unfamiliar with Wikipedia policy and convention, and so their behavior may initially appear to be disruptive. Responding to these new users with excessive force can discourage them from editing in the future (see Do not bite the newcomers).

Protection
A user may be blocked when necessary to protect the rights, property, or safety of the Wikimedia Foundation, its users, or the public. A block for protection may be necessary in response to:
 * persistent or severe personal attacks;
 * personal, professional, or legal threats (including outside the Wikipedia site);
 * actions placing users in danger;
 * actions that may compromise the safety of children, in accordance with Child protection;
 * disclosures of others' personal information (whether or not the information is accurate);
 * persistent copyright violations;
 * persistent posts of unreferenced, poorly or incorrectly referenced, or potentially defamatory information about living persons; or
 * an account appearing to have been compromised (as an emergency measure), i.e. there is some reason to believe the account is being used by someone other than the person who registered the account.

When blocking in response to personal information disclosures or actions that place users in danger, consider notifying the Arbitration Committee by e-mail (arbcom-en@wikimedia.org) about the disclosure or danger, as well as contacting someone with oversight permissions to request deletion of the material in question.

Disruption
A user may be blocked when their conduct severely disrupts the project; that is, when their conduct is inconsistent with a civil, collegial atmosphere and interferes with the process of editors working together harmoniously to create an encyclopedia. A block for disruption may be necessary in response to:
 * vandalism;
 * gross incivility;
 * harassment;
 * spamming;
 * deliberately tripping the edit filter
 * breaching the policies or guidelines, especially sock puppetry;
 * attempts to coerce actions of editors through threats of actions outside the Wikipedia processes, whether onsite or offsite.

Edit warring, especially breaches of the three-revert rule, often results in a block, either from the pages the user is disrupting or from the entire site.

Disruption-only
Some types of user accounts are considered disruptive and may be blocked without warning, usually indefinitely:
 * Accounts used exclusively for disruptive purposes, such as vandalism.
 * Accounts that appear, based on their edit history, to exist for the sole or primary purpose of promoting a person, company, product, service, or organization. See Conflict of interest and Spam.
 * Accounts with inappropriate usernames.
 * Public accounts (where the password is publicly available or shared with a large group).
 * Bots operating without approval or outside their approval, or that appear to be malfunctioning.

Open or anonymous proxies
Open or anonymous proxies may be blocked on sight.

Non-static IP addresses or hosts that are otherwise not permanent proxies typically warrant blocking for a shorter period of time, as the IP address is likely to be reassigned, or the open proxy is likely to be closed. Many Tor proxies, in particular, are "exit nodes" for only a short time; in general, these proxies should not be blocked indefinitely without consideration. See Blocking IP addresses for further details.

There is also a Wikipedia project, the WikiProject on open proxies, which seeks to identify and block open proxy servers.

Enforcing bans
A Wikipedia ban is a formal revocation of editing privileges on all or part of Wikipedia. A ban may be temporary and of fixed duration, or indefinite and potentially permanent.

Blocks may be imposed as a technical measure to enforce a ban. Such blocks are based on the particulars of the ban. Bans that apply to all of Wikipedia—that is, they are not partial—may be backed up by a sitewide block, which is usually set to apply for the period of the ban. Other bans may be enforced with a partial block.

"Not here to build an encyclopedia"
This often-used blocking rationale is described at.

Evasion and enforcement
An administrator may reset the block of a user who intentionally evades a block, and may extend the duration of the block if the user engages in further blockable behavior while evading the block. User accounts or IP addresses used to evade a block should also be blocked.

Administrator conflicts and involvement
Administrators must not block users with whom they are engaged in a content dispute; instead, they should report the problem to other administrators. Administrators should also be aware of potential conflicts involving pages or subject areas with which they are involved. It is acceptable for an administrator to block someone who has been engaging in clear-cut vandalism in that administrator's userspace.

Cool-down blocks
Blocks intended solely to "cool down" an angry user should not be used, as they often have the opposite effect. However, if an angry user is also being disruptive, the user can be blocked to prevent further disruption.

Recording in the block log
Blocks should not be used solely for the purpose of recording warnings or other negative events in a user's block log. The practice, typically involving very short blocks, is often seen as punitive and humiliating.

Very short blocks may be used to record, for example, an apology or acknowledgement of mistake in the block log in the event of a wrongful or accidental block, if the original block has expired. (If it has not, the message may be recorded in the unblocking reason.)

Against the blocking administrator
A blocked administrator can block the blocking administrator, but should only do so in exceptional circumstances where there is a clear and immediate need, such as in the case of a compromised account. Use of the block tool to further a dispute or retaliate against the original blocking administrator is not allowed. If in doubt, report the issue on the Administrators' noticeboard for incidents.

Requesting blocks
Disruptive behavior can be reported, and blocks requested at a specialized venue such as Administrator intervention against vandalism or, if appropriate, Administrators' noticeboard/Incidents. Users requesting blocks should supply credible evidence of the circumstances warranting a block. Administrators are never obliged to place a block, and are free to investigate the situation for themselves. Prior to imposing a block, administrators are expected to be fully familiar with the circumstances of the situation. See also.

Dealing with off-wiki block requests
Administrators who use Wikipedia-related IRC channels and Discord are reminded that, while these channels have legitimate purposes, discussing an issue on IRC or Discord necessarily excludes those editors who do not use IRC or Discord from the discussion (and excludes all non-administrators from the discussion if it takes place in an admin-only channel such as #wikipedia-en-admins), and therefore, such discussion is never the equivalent of on-wiki discussion or dispute resolution. Consensus about blocks or other subjects should not be formed off-wiki.

As the practice of off-wiki "block-shopping" is strongly discouraged, and that except where there is an urgent situation and no reasonable administrator could disagree with an immediate block (e.g. ongoing vandalism or serious violations of the policy on biographies of living persons), the appropriate response for an administrator asked on IRC or Discord to block an editor is to refer the requester to the appropriate on-wiki noticeboard.

Self-requested blocks
Sometimes, people request that their account be blocked, for example to enforce a wikibreak. There is a category of administrators who will consider such requests. As an alternative to requesting a self-block, users may use the Wikibreak Enforcer, a user script that can prevent a user from logging in.

Preliminary: education and warnings

 * Some of the key precepts of this section may be explained using Before blocking.

Before a block is imposed, efforts should be made to educate users about Wikipedia policies and guidelines, and to warn them when their behavior conflicts with these. Welcome newcomers, do not bite them, and assume that most people who work on the project are trying to help it, not hurt it. Newcomers should make an effort to learn about our policies and guidelines so that they can learn how to avoid making mistakes. A variety of template messages exist for convenience, although purpose-written messages are often preferable. Template warnings that state that a user may be blocked for disruption or other blockable behavior may also be issued by regular editors rather than by administrators only.

However, warnings are not a prerequisite for blocking. In general, administrators should ensure that users who are acting in good faith are aware of policies and are given reasonable opportunity to adjust their behavior before blocking, and it may be particularly desirable to communicate first with such users before blocking. On the other hand, users acting in bad faith, whose main or only use is forbidden activity (sockpuppetry, vandalism, and so on), do not require any warning and may be blocked immediately.

Explanation of blocks
Blocking is a serious matter. The community expects that blocks will be made for good reasons only, based upon reviewable evidence and reasonable judgment, and that all factors that support a block are subject to independent peer review if requested.

Notifying the blocked user
Administrators must supply a clear and specific reason why a user was blocked. Block reasons should avoid the use of jargon as much as possible so that blocked users may better understand them. Administrators should notify users when blocking them by leaving a message on their user talk page. It is usually easier to explain the reason for a block at the time it is applied than afterwards.

When implementing a block, a number of pro forma block reasons are available in a drop-down menu; other or additional reasons can also be added. Users can be notified of blocks and block reasons using a number of convenient template messages—see Category:User block templates and.

Other important information
If there are any specific recommendations or circumstances that a reviewing administrator would need to know, or that may help to avoid administrator disputes upon review of a block, the blocking administrator should consider including this information in the block notice. For example:
 * When there is information or evidence that may not be obvious, may not be fully appreciated, or may otherwise be relevant.
 * Prior endorsement that if any administrator wishes to unblock, or there is consensus for it, they may without consulting the blocking administrator.
 * Suggested conditions for an unblock.

Confidential evidence
If the rationale for a block depends on information that is not available to all administrators, that information should be sent to the Arbitration Committee, a Checkuser, or an Oversighter (as applicable) for action. These editors serve as functionaries; they are qualified and trusted to handle non-public evidence, and they operate under strict controls. The community has rejected the idea of individual administrators acting on evidence that cannot be peer-reviewed. Administrators must be able to justify their blocks using evidence visible on Wikipedia, even if it includes aspects only accessible by other administrators (e.g. revisions or log details that are redacted, and deleted pages).

Administrators who are also Checkusers or Oversighters may block users based on non-public information either revealed through the checkuser function page, or revisions and log details that have been suppressed ("oversighted"). These administrators may also make blocks based on off-wiki evidence of sockpuppetry and undeclared paid editing submitted through the email addresses below. This evidence is inaccessible to administrators. As such, an administrative action is generally viewed to be made in the user's capacity as a Checkuser or Oversighter, although the action itself is an administrative one. All such blocks are subject to review by other members of the functionary team, and direct review by the Arbitration Committee.
 * Contact details: individual Checkusers and Oversighters are listed on the relevant pages. Private evidence involving undisclosed paid editing may be sent to '. Other matters requiring Checkuser attention may be sent to '.

Implementing blocks
Technical instructions on how to block and unblock, and information on the blocking interface, are available at mw:Help:Blocking users. The following is advice specifically related to blocking and unblocking on Wikipedia.

IP address blocks
In addition to the further advice, there are special considerations to take into account when blocking IP addresses. IP address blocks can affect many users, and IP addresses can change. Users intending to block an IP address should at a minimum check for usage of that address, and consider duration carefully. IP addresses should rarely, if ever, be blocked indefinitely. You should notify the Wikimedia Foundation if the IP is related to a sensitive organization or a government agency.

Collateral damage
A block of a range of IP addresses may unintentionally affect other users in that range. Before blocking an IP range, especially for a significant time, you should check for other users who may be unintentionally affected by the range block: If any are found, an IP block exemption ensures they will not be affected.
 * unregistered users – Range Contributions – X!'s tools
 * registered users – ask a user with checkuser access to check.

Duration of blocks
The purpose of blocking is prevention, not punishment. The duration of blocks should thus be related to the likelihood of a user repeating inappropriate behavior. Longer blocks for repeated and high levels of disruption are to reduce administrative burden; they are made under the presumption that such users are likely to cause frequent disruption or harm in future. Administrators should consider:
 * the severity of the behavior;
 * whether the user has engaged in that behavior before.

Blocks on shared or dynamic IP addresses are typically shorter than blocks on registered accounts or static IP addresses made in otherwise similar circumstances, to limit side-effects on other users sharing that IP address.

While the duration of a block should vary with the circumstances, there are some broad standards:
 * incidents of disruptive behavior typically result in blocks of from a day to a few days, longer for persistent violations;
 * accounts used exclusively for disruption may be blocked indefinitely without warning;
 * protective blocks typically last as long as protection is necessary, often indefinitely.

Indefinite blocks
An indefinite block is a block that does not have a definite (or fixed) duration. Indefinite blocks are usually applied when there is significant disruption or threats of disruption, or major breaches of policy. In such cases, an open-ended (indefinite) block may be appropriate to prevent further problems until the matter can be resolved by discussion. As with all blocks, it is not a punishment. It is designed to prevent further disruption, and the desired outcome is a commitment to observe Wikipedia's policies and guidelines, and to stop problematic conduct in the future.

Indefinite does not mean "infinite" or "permanent"; it just means that no automatic expiration time (or duration) for the block has been set. An indefinitely blocked user may later be unblocked in appropriate circumstances. In particularly serious cases in which no administrator would be willing to lift the block, the user is effectively banned by the community.

Block log
If the block arose from a discussion per, please include a link to the discussion in the block log. If the block is enforcing a community sanction, please note this. If consensus was to allow for regular administrative review rather than requiring community review, per, that should be noted in the log as well.

Setting block options
Several options are available to modify the effect of blocks, which should be used in certain circumstances:

Editing block options

 * Sitewide block will prevent the user from editing any page on Wikipedia with the exception of their own user talk page. This is the option that is set by default, and should be used when there is a reasonable assumption that the account would disrupt any page, such as vandalism-only accounts or users that are clearly not here to write an encyclopedia.
 * Partial block will prevent the user from editing a specific set of pages, or from a particular set of namespaces. Either option may be set, or a combination of both may be chosen. There is a software limit of 10 pages per block; beyond this, sitewide blocking should be considered instead.

Standard block options

 * Autoblock any IP addresses used will apply an autoblock, or automatic block, on the IP address that the account was last using, as well as any subsequent IP addresses the account tries to edit from while they are blocked with this option set. If a different non-IP block exempt user account logs in from an autoblocked IP address and tries to edit, the user account will also be added to the autoblock list. This option should typically be disabled when blocking unapproved or malfunctioning bots (so as not to block the bot's operator or any other bots using that IP address), though it should be enabled when blocking accounts for disruptive or malicious behavior. This option is enabled by default and is only available when applying a block to an account.
 * Prevent account creation will restrict the user from accessing the Special:CreateAccount function page (and hence restricts the user from creating new accounts) for the duration of the block. If applied to a hard-blocked IP address or range, it will also prevent all user accounts who are not IP block-exempt from being able to create additional accounts if they attempt to do so while behind the blocked IP address or range. If applied to a user account and with the autoblock option also set, it will also prevent accounts from being created on the IP address that the account was last using. It should typically be disabled when blocking accounts with inappropriate usernames (to allow the user to create a new account with an appropriate one), though it should be enabled when blocking bad-faith usernames (e.g. clearly threatening, abusive, or clear attacks toward other editors) or vandalism-only accounts.
 * Prevent user from sending email will restrict the user from accessing the Special:EmailUser function page (and hence restricts the user from sending any emails to user accounts) for the duration of the block. This option is not checked by default and should not be enabled when blocking an account except only in cases where either the blocked user abuses it, or uses it in order to harass, threaten, intimidate, or cause disruption toward other editors. In instances when administrators feel that email abuse is extremely likely, they may use their discretion and enable this option to prevent it from occurring. When enabled, efforts should be taken to ensure that the user's talk page remains unprotected and that the user is aware of other avenues (such as the Unblock Ticket Request System) through which they can discuss the block. When applied to a hard-blocked IP address or range, it will also prevent all user accounts who are not IP block exempt from being able to email other accounts if they attempt to do so while behind the blocked IP address or range. If applied to a user account and with the autoblock option also set, it will not have a direct effect on the IP address that the account was last using, since IP address users do not have access to the Special:EmailUser function page.


 * Prevent this user from editing their own talk page while blocked, if checked, will prevent the blocked user from editing their own user talk page (and hence, the ability for them to create unblock requests) during the duration of their block. This option is not checked by default, and typically should not be checked; editing of the user's talk page should be disabled only in cases of continued abuse of their user talk page, or when the user has engaged in serious threats, accusations, or attempts at outing that must be prevented from re-occurring. The protection policy has further details in cases where other users are repeatedly causing disruption to the user talk page of a blocked user.
 * Apply block to logged-in users from this IP address will disallow all non IP-block exempt user accounts from editing from the IP address or range during the duration of the block. If the ability to create accounts or send email to other users is also disallowed, these functions will also be disallowed for any non-exempt user accounts who are attempting to do so behind the blocked IP address or range. This option should typically not be checked, and is typically only used in cases of long-term abuse, sock puppetry, for IP addresses with a history of significant and high level abuse, or for being an open proxy or location host. See hard block under the IP address common block list below. This option is disabled by default and is only available when applying a block to an IP address or IP range.

Common blocks imposed
There are two common blocks that may be imposed on registered accounts:


 * A soft account block (autoblock disabled, account creation allowed) will only block the specific account from editing. An autoblock is not applied to the IP address the account last used, and other accounts that log in from the IP address are allowed to edit as normal. This is generally used in situations such as blocking promotional usernames or to enforce other username policy violations. This allows the blocked account to create a new account with a username that is in compliance with the username policy, or to simply choose to edit anonymously instead.
 * A hard account block (autoblock enabled, account creation disabled) will apply an autoblock to the IP address the account last used to edit, and disable the ability for the user to create other accounts during the duration of the block. Any additional IP address(es) that the account attempts to edit from during the duration of the block is also automatically blocked and added to the autoblock list. Any non-IP block exempt accounts that attempt to edit from an autoblocked IP address will not be able to do so, and will also be automatically blocked and added to the autoblock list. Accounts also cannot be created by any autoblocked IP address(es), or by any non-IP exempt accounts while logged in behind an autoblocked IP address. This is typically used in cases of blocking vandalism or to prevent other disruption.

There are two common blocks that may be imposed on IP addresses or ranges:


 * A soft IP address block (anon. only, account creation blocked) is used in most cases of disruption – including vandalism and edit warring, and prevents only anonymous users from editing. It also restricts any account creation by the IP address or by any non-IP exempt user accounts while behind the blocked IP address. Allowing account creation from a blocked IP address or range is rare and performed only in special circumstances.
 * A hard IP address block (account creation blocked, apply block to logged-in users from this IP address) disables all editing and account creation from behind the blocked IP address, whether they be attempted anonymously or using an account (with the exception of accounts that are IP-block exempt). This is typically used when the level of vandalism or disruption via creation of "throwaway" accounts is such that all editing from the IP address is to be prevented except after individual checking of requests. Open proxies are hard-blocked on detection, and Tor IP addresses are automatically blocked by the Tor block extension.

Blocking bots
Automated or semi-automated bots may occasionally not operate as intended for a variety of reasons. Bots (or their associated IP address should the actual bot not be readily identifiable) may be blocked until the issue is resolved. Bots should be softblocked (autoblock disabled) to ensure the autoblock doesn't affect other unrelated bots sharing the same IP. If only a single task is malfunctioning and the bot supports disabling individual tasks, it is preferable to disable the single malfunctioning task so that other bot tasks can continue running.

Bots that are unapproved, or usernames that violate the username policy due to a resemblance to a bot, are immediately and indefinitely blocked if they violate the bot policy, most commonly by editing outside the operator's or their own userspace.

The edits of a bot are considered to be, by extension, the edits of the editor responsible for the bot. As a result, should a bot operator be blocked, any bot attributed to them may also be blocked for the same duration as that of the blocked editor.

Recording in the block log after a "clean start"
Editors may cite "clean start" and rename themselves, asking that their previous username not be disclosed. If such editors have been blocked previously, the administrator who has been requested to make the deletion should contact a Checkuser so that the connection between the accounts can be verified. The Checkuser should then consider adding short blocks to the new account to denote each entry in the user's old account log. Such short blocks should provide protection in case the "clean start" was based on a genuine risk of off-wiki harassment, by not disclosing the previous username, while at the same time eliminating the possibility of avoiding the scrutiny of the community.

The short blocks should be described in the block summary as "previous account block" and the final duration of the block should be noted. Blocks placed in error and lifted early should not be noted at all.

Unblocking
Unblocking or shortening of a block is most common when a blocked user appeals a block. An uninvolved administrator acting independently reviews the circumstances of the block, the editor's prior conduct, and other relevant evidence, along with any additional information provided by the user and others, to determine if the unblock request should be accepted. Common reasons include: the circumstances have changed, a commitment to change is given, the administrator was not fully familiar with the circumstances prior to blocking, or there was a clear mistake.

Unacceptable unblocking
Unblocking will almost never be acceptable: Each of these may lead to sanctions for misuse of administrative tools—possibly including removing administrator rights—even for first-time incidents.
 * When it would constitute wheel warring.
 * To unblock any of one's own accounts, except in the case of self-imposed blocks.
 * When the block is implementing a community sanction that has not been successfully appealed. The community may choose to allow a block to be reviewed in the normal way, by consulting with the closing/blocking administrator, rather than requiring a formal appeal to the community. If there is consensus to allow this, it shall be noted in the closing statement and block log.
 * When the block is designated as a CheckUser or Oversight block, and the unblocking administrator is not a member of the designated functionary group or does not have approval from someone in that group to carry out the action.
 * When the block is explicitly enforcing an active Arbitration remedy. Arbitration enforcement blocks may be appealed using the special appeal provisions.

There is no predefined limit to the number of unblock requests that a user may issue. However, disruptive use of the unblock template may prompt an administrator to remove the blocked user's ability to edit their talk page. In this case, a block may still be appealed by submitting a request to the Unblock Ticket Request System.

Unblock requests
As part of an unblock request, uninvolved editors may discuss the block, and the blocking administrator is often asked to review or discuss the block, or provide further information. Since the purpose of an unblock request is to obtain review from a third party, the administrators who performed the block should not decline unblock requests. Also, by convention, administrators don't usually review more than one unblock request regarding the same block.

Except in cases of unambiguous error or significant change in circumstances dealing with the reason for blocking, administrators should avoid unblocking users without first attempting to contact the blocking administrator to discuss the matter. If the blocking administrator is not available, or if the administrators cannot come to an agreement, then a discussion at Administrators' noticeboard is recommended.

Administrators reviewing a block should consider that some historical context may not be immediately obvious. Cases involving sockpuppets, harassment, or privacy concerns are particularly difficult to judge. At times such issues have led to contentious unblocks. Where an uninformed unblock may be problematic, the blocking administrator may also wish to note as part of the block notice that there are specific circumstances, and that a reviewing administrator should not unblock without discussing the case with the blocking admin (or possibly ArbCom) to fully understand the matter.

If users claim they wish to contribute constructively but there are doubts as to their sincerity, the 2nd chance template can be used to allow them to demonstrate how they will contribute to the encyclopedia, should their unblock request be granted.

Any user may comment on an unblock request; however, only administrators may resolve the request (by either declining or unblocking).

Blocks in temporary circumstances
Some types of blocks are used in response to particular temporary circumstances, and should be undone once the circumstance no longer applies:
 * Blocks on open or anonymous proxies should be undone once it is confirmed that they have been closed (but be aware some open proxies may be open only at certain times, so careful checking may be needed that it really is apparently no longer in use that way).
 * Blocks of unapproved or malfunctioning bots should be undone once the bots gain approval or are repaired.
 * Blocks for making legal threats should be undone once the threats are confirmed as permanently withdrawn and no longer outstanding.

Unblocks in temporary circumstances
Users may be temporarily and conditionally unblocked to respond to a discussion regarding the circumstances of their block. Such temporary and conditional unblocks are made on the understanding that the users may not edit any pages (besides their user talk page) except the relevant discussion page(s) explicitly specified by the unblocking admin. The users are effectively banned from editing any other pages, and breaching this ban will be sanctioned appropriately. When the discussion concludes, the block should be reinstated unless there is a consensus to overturn the block.

CheckUser blocks
Without first consulting with a CheckUser and receiving approval, administrators must not undo or "loosen" (lowering the block duration or disabling any block options applied) any block set, or unblock any user when the block is specifically identified as a "CheckUser" block, such as through the use of the checkuserblock or checkuserblock-account templates in the action summary, or by explicitly stating in the action summary that the block is a "CheckUser block" or is "checkuser-related". If an administrator believes that a CheckUser block has been made in error, the administrator should first discuss the matter with the CheckUser in question, and if a satisfactory resolution is not reached, should e-mail the Arbitration Committee. Any unauthorized reversal or alteration of such a block and without prior consultation and approval from a CheckUser may result in removal of permissions.

Oversight blocks
Without first consulting with an Oversighter and receiving approval, administrators must not undo or "loosen" (lowering the block duration or disabling any block options applied) any block set, or unblock any user when the block is specifically identified as an "Oversight" block, such as through the use of the OversightBlock template in the action summary, or by explicitly stating in the action summary that the block is an "Oversight block" or is "Oversight-related". Appeals of any Oversight blocks that are set must be sent to either the Oversight team via email to be discussed in private by the English Wikipedia Oversight team, or to the Arbitration Committee. Blocks may still be marked by the blocking Oversighter as appealable only to the Arbitration Committee, per the 2010 statement, in which case appeals must only be directed to the Arbitration Committee. Any unauthorized reversal or alteration of such a block and without prior consultation and approval from an Oversighter may result in removal of permissions.

Conditional unblock
Administrators may, with the agreement of the blocked user, impose conditions when unblocking. Unblock conditions are designed to prevent recurrence of the behaviour that led to the block (such as a page ban to prevent further edit warring).
 * If the blocked user does not reach an agreement on proposed unblock conditions with an administrator, the blocked user may post another block appeal.
 * Administrators have discretion to set the expiry of unblock conditions, provided that:
 * The unblock conditions of blocks that expire after one year or less will expire after no more than a year,
 * The unblock conditions of blocks that expire after more than a year (including indefinite) may expire up to and including indefinitely.
 * Unblock conditions may include page bans, topic bans, interaction bans, revert restrictions, single account restrictions and other restrictions at the discretion of the unblocking administrator.
 * A partial block may be used to enforce the unblock conditions of a sitewide block.
 * If editors breach the unblock conditions or engage in fresh misconduct, they may be blocked or further restricted.
 * After the blocked user has accepted the conditions and been unblocked, the conditions may be appealed only to the unblocking administrator or to Administrators' noticeboard.
 * The user will be notified of unblock conditions on their talk page when they are unblocked and a diff/permalink containing the restrictions must be included in the unblock log rationale.
 * The restriction should (but do not have to be) be logged at Final warnings / Unblock conditions.

Partial blocks
Partial blocks may be used at the discretion of any administrator in accord with the rest of the blocking policy, or community consensus. They may also be used to enforce editing restrictions or as a requirement for conditional unblocks.

The affected editor may request an unblock following the procedures listed in, using the unblock template, or appealing at the Administrators' noticeboard. Administrators can unblock a user when they feel the block is unwarranted or no longer appropriate, in accordance with the blocking policy.

Global blocks
Global blocking is a MediaWiki extension available to stewards to prevent cross-wiki disruption from an IP address or a range of IP addresses. When an IP address or range of IP addresses is globally blocked, they are prevented from editing any public Wikimedia wiki, except for Meta-Wiki, where globally blocked users may appeal the decision. (A global block is not the same as a global ban.) When a user's editing is prevented by a global block, the contents of MediaWiki:Wikimedia-globalblocking-ipblocked (formerly MediaWiki:Globalblocking-ipblocked) are shown as an error message (analogous to MediaWiki:Blockedtext for locally blocked users). Registered users cannot be globally blocked. The analogous action is global locking, which prevents anyone from logging into the account.

A current list of globally blocked IP addresses is available at Special:GlobalBlockList.

Unblocking and appeal
Local whitelisting — An IP address which is globally blocked can be unblocked locally (to edit the specific wiki concerned only), by any local administrator, at Special:GlobalBlockWhitelist. It is not possible to override global locks locally.

Appeal against a global block — Globally blocked IP addresses and globally locked users may appeal through the email queue to. Globally blocked IP addresses may also appeal through their meta talk page, if access to it has not been revoked.