Requests for comment/Global ban requirements

Add links
From Meta, a Wikimedia project coordination wiki

This is a subpage; for more information, see the Requests for comments page.

Hello everyone!

This RfC is inspired by Maggie Dennis's email on Wikimedia-l describing some potential issues with our current requirements for proposing global bans.

The existing criteria for global bans emphasizes the cross-wiki nature of global bans, specifying a requirement for cross-wiki abuse and that the user be indefinitely blocked or banned on two or more projects.

Given the changing landscape of Wikimedia since the policy's last update (current wording of the important bits remains unchanged since 2011), with the proliferation of affiliates and off-project spaces, and rising awareness of the importance of combating disinformation and other organized attempts to undermine our movement's priorities, it seems pertinent to gauge current community views on the topic.

Historically, community global bans have shifted from relatively straightforward long-term abuse type cases (that are now generally handled with locks) to more complex cases, often involving subtle sockpuppetry, disinformation, off-wiki actions, and harassment. Good examples of these more modern cases include the successful ban requests of James Salsman, Kubura, and Til Eulenspiegel.

Though these three cases did meet the cross-wiki criteria, as they were blocked on two or more projects, they focused primarily on their incredibly problematic activities on one project, especially with Kubura and Til Eulenspiegel.

Had there not been these additional blocks cross-project, the arguments for their ban (and the harm they inflicted) would be next to unchanged, but the ban would have been procedurally invalid. In other words, regardless of the damage to a given local community, unless it spreads elsewhere, the global community is unable to pursue global action through this mechanism.

Certainly, maintaining the criteria of being blocked on 2 or more projects does help weed out a lot of cases that will clearly go nowhere. However, it has the potential to make legitimate global ban requests inert. It might be beneficial to loosen up the criteria for proposing a global ban, maintaining a focus on cross-wiki abuse, but adding content about damage to the Wikimedia movement and mission. This would loop in the editors engaged in harmful disinformation campaigns, exclusionary bigotry, harassment/hounding, and other issues caused by people who do not happen to be blocked on multiple projects.

To note, this would not make it easier to enact global bans, just to propose them. It would give the global community more room in determining whether a user should be subject to a global ban, which is especially helpful in light of the changing landscape of problematic activities we find on Wikimedia projects.

Thank you for your time, and happy editing! Vermont 🐿️ (talk) 18:54, 1 June 2022 (UTC)[reply]

Discussion[edit]

It seems to be a reasonable proposal in principle. · · · Peter (Southwood) (talk): 16:38, 2 June 2022 (UTC)[reply]

  • I'm not keen on using the meta-wiki global ban process on a user even in cases where no local project supports banning or even blocking the contributor, especially so if the user in question is a single-project contributor. Perhaps requiring 2+ bans isn't needed, but we shouldn't need the "global community" to override a local project that doesn't have a problem with a specific contributor. Am I missing something? — xaosflux Talk 18:22, 2 June 2022 (UTC)[reply]
    The issue, in my view, is primarily the assumption of current policy that being blocked on multiple projects is the only good indicator that a person has engaged in harm necessitating a ban, which carry’s a few outdated assumptions within it: that these projects have functioning enforcement processes to handle problematic users, that damage on one project is always limited to that project, that global bans are solely for on-wiki activities (I.e. not in all movement spaces), and fundamentally the incorrect assumption that a user cannot have harmed the movement without being blocked on multiple projects.
    My goal with this isn’t to override local communities, and I don’t believe that would be the result any more than it currently is. (Note mandatory notifications on all projects the user edits) My point is primarily that current policy may prevent the global ban requests of users who have irreparably harmed the movement without meeting the criteria of being blocked in two or more projects. This means that those requests can go to discussions without being immediately removed. If it isn’t a good request regardless, it can just be closed as usual. Vermont 🐿️ (talk) 03:05, 4 June 2022 (UTC)[reply]
    It's not that hard to get indef-blocked on at least 1 project (as GS's can do those on S/M projects alreday) - so maybe we don't require 2? Not sure how much utility this on-wiki process should have for dealing with off-wiki activities. UCOC and Foundation bans may be better suited in those cases? — xaosflux Talk 17:50, 4 June 2022 (UTC)[reply]
    Even then, if you are only causing disruption to one community then this isn't really needed either. I'd be fine extending the "2 places" to include non-project official areas, such as phabricator, affiliate projects, in-person events, etc - but don't think if someone is just posting over on reddit the community ban process is the best. — xaosflux Talk 22:56, 6 June 2022 (UTC)[reply]
    @Xaosflux, I think here the issue is can not be left with foundation bans, as we have seen after enwiki admin's fram case the foundation has refused to override the local communities. To bring some solid argument I would like to give example from two cases. One at mrwikipedia project of long term harassment and stalking of women editors(By the way the user is blocked on three major projects!). And persistent copyvio by one editor for last ten years. Now, however they both might be lockable(if there is such a word!). Despite of worsening situations at our projects. No admin/crat has ever taken cognizance of it. Neither they paid any attention to the reports they received on village pump. I can provide you the diffs for all of this. But no local consensus was ever achieved on this, nor anyone from outside were willing to take any actions as there is a local crat and admin available. But I can see local community was ignored in Til Eulenspiegel's case and I am sure in many other cases. Owing to their not so good behavior on several scales. So, I think when their behavior on many other projects is screaming badly why do we stop at the threshold of overriding local communities? QueerEcofeminist [they/them/their] 02:32, 15 June 2022 (UTC)[reply]
    Hi, so for your user with multiple project blocks - removing project blocks as a requirement isn't an issue right? For your user that has no local project blocks, on any projects that have communities, you want volunteers here to override those local projects decision to not block that user and block them anyway? That's what I'm not loving about this scenario. A global ban is every project, and the foundation doesn't seem to have any problem giving those out (the situation you referred to on enwiki was when the foundation wanted to ban someone from only one project) under their existing framework, or under the UCOC process. — xaosflux Talk 09:53, 15 June 2022 (UTC)[reply]
  • In my opinion Global locks meeds a larger overhaul (which I have said many times before) - which also affects when global bans will be applicable. There are too many cases that users locked either does not prevent new disruption (as they are inactive and/or already blocked in wikis that causes problem), or a bit controversial. By the way, please notice not every wikis have a community ban process. Such a process does not exists at Chinese Wikipedia at all and we consider such a process evil.--GZWDer (talk) 18:31, 2 June 2022 (UTC)[reply]
@GZWDer: Hello, what would I read if I wanted to understand exactly the difference between the CN and EN blocking policies? ~^\\\.rT'{~ g 04:40, 24 June 2022 (UTC)[reply]
What I mean: in some wikis there is a banning policy (not to be confused with blocking), but in some other ones there are not.--GZWDer (talk) 06:58, 24 June 2022 (UTC)[reply]
So, if there was vandalism or pure abuse, obviously an administrator warns and blocks. But is that the only form of dispute that is dealt with by administrators and blocking? What about when editors simply argue for weeks, but do not really damage the content or the output of the encyclopaedia, are they just allowed to continue arguing on condition that the content is safe? On the en wiki, a person can be banned for being difficult to understand, or for not being agreed with, even if their article edits are generally helpful. You must fear the other editors to a certain extent because if five or ten of them simply do not like you and want rid of you, you have to hide from all arguments or you will be banned forever. It's kind of sad really, in my opinion. ~^\\\.rT'{~ g 21:25, 24 June 2022 (UTC)[reply]
I think this is a very valuable point you make, User:RTG. Sometimes the criteria used on smaller wikimedia projects to ban a contributor can be opaque at best. Only a few weeks ago, on one of the smaller wikimedia projects, a contributor with only a handful of contributions (all constructive!) was banned without even nominating them for a ban (so no defence possible) because they were from Australia and had Telstra as ISP (as do many Australians!). Earlier this week, on the same wikimedia project, a contributor was nominated for a ban because they created a new article, and another contributor (with a questionable editing record) didn't like "how they sounded". The smaller the wikimedia project, the less oversight there tends to be to curb this kind of undemocratic behaviour. Using number of bans on individual projects as a requirement for global ban nomination is therefore very dangerous. ArticCynda (talk) —Preceding undated comment added 14:16, 3 July 2022 (UTC).[reply]
It's not truly public or content focused is it, and like a lot of things the longer it goes on the more difficult it is to do things differently. It's like a cop-out of getting things to work the best. ~^\\\.rT'{~ g 13:11, 19 July 2022 (UTC)[reply]
@ArticCynda: If your recounting of the events is truthful then it is very concerning. I think you should file a RfC on how that project is inappropriately handling bans, or more generally a RfC on how small wikis are susceptible to these problems. Dsuke1998AEOS (talk) 02:57, 12 December 2022 (UTC)[reply]
@Dsuke1998AEOS: Looking at the RfCs already open, it's quite clear to me that abuse of admin powers on smaller wikis is much more widespread than we would assume - maybe only the tip of the iceberg is reported here. The inappropriate handling of bans is only one symptom of a culture where abuse of powers by admin became normalised. It should be noted that many small wikis sometimes only have a handful of regular contributors, so abuse patters may go unnoticed. Unfortunately, I don't have a good solution to this problem. ArticCynda (talk) 20:25, 28 November 2023 (UTC)[reply]
  • I think my feelings align somewhat with Xaosflux on this. In addition, I wonder if this new scenario isn't better off being handled by the forthcoming U4C rather than trying to adjust the community method for a global ban. Best, Barkeep49 (talk) 22:44, 6 June 2022 (UTC)[reply]
  • Support Support There's a user that I know of, who I won't name (email me if you want to know) who has repeatedly harassed other users, on several projects, but is only indeffed on one wiki, and a lot of LTAs who've been banned for harassment aren't necessarily indeffed on two or more projects. --SHB2000 (talk | contribs) 11:01, 7 June 2022 (UTC)[reply]
    @SHB2000 regarding your example: on all the projects this person isn't blocked on, given that they are actually harassing people on those projects, why are those projects not stopping that contributor? — xaosflux Talk 13:44, 7 June 2022 (UTC)[reply]
The blocks are counted per user. So, if the main account is blocked on one wiki and a sock on another, this will qualify as the required two blocks. Ruslik (talk) 19:44, 9 June 2022 (UTC)[reply]
I think an block on one wiki and both an sock and no live edits on another would classify as two blocks. No live edits would follow the same criteria as xtools uses, users with no edits on the second wiki would never fullfill this requirement.--Snævar (talk) 22:42, 9 June 2022 (UTC)[reply]
Apologies @Xaosflux for a little late reply. Unfortunately, this is a user who rather sneakily cribs and complains on other wikis about how reverting their edits (while block evading) are "political censorship" (English translation of censure politique) or "politically-inspired vandalism" (English translation of Politiek geïnspireerd vandalisme) and the wikis they complain to often take their word for it hence why the other projects aren't doing anything. Chances are, they're probably stalking my contributions here, but email me if you want to know who that contributor is. SHB2000 (talk | contribs) 05:02, 27 June 2022 (UTC)[reply]
  •  Weak support I wouldn't be opposed to this, but what I can say is that these types of RfC could use a bit of improvement to up its quality. --DarkMatterMan4500 (talk) (contribs) 16:50, 13 June 2022 (UTC)[reply]
  • I think this needs a certain limitation on the type of blocks that can trigger this. If there's only one block, it needs to be for something really bad to allow us to prejudge how the user would fare editing other projects. --Rschen7754 18:03, 15 June 2022 (UTC)[reply]
  • Neutral. We should build for soft security, including at the cross-project scale. 1) sometimes it is clearly reasonable to apply a cross-project ban, sometimes it isn't. [we already do this for spam and abuse, no?] so our global policies shouldn't suggest this is impossible. 2) Focus on making it easier to flag the activity of a user for cross-project attention, even if it's not clear that a community ban is likely. –SJ talk  23:33, 15 June 2022 (UTC)[reply]
  • I will Support Support lowering it down to 1 block. I think it makes sense considering how I have seen few cases where there is abuse across several wikis which is not at the level of triggering block, but have impact on participants of those communities nonetheless. I think it should also be clarified on requests that the nature of abuse is crosswiki, or if a local ban isn't preventing the abuse. I just don't think that defining crosswiki abuse by 'x number of block' and setting it as a criteria is good idea.--BRP ever 03:09, 18 June 2022 (UTC)[reply]
    Saidly @BRPever that's why I Oppose Oppose this RFC (wow, I'm the first ever user who voted oppose), Indef blocked only on one wiki may also be an alert for us that that blocking wiki might be a caesaropapism wiki, just see Special:Centralauth/Cekli829 for example, a trustful user within Azerbaijani-speaking users, has blocked by Armenian for years just because of the worse-and-worse AZE-ARM relations. Liuxinyu970226 (talk) 09:50, 22 June 2022 (UTC)[reply]
    There's no global account for Cekli2011. NguoiDungKhongDinhDanh 09:53, 22 June 2022 (UTC)[reply]
    @NguoiDungKhongDinhDanh: Fixed typo. --Liuxinyu970226 (talk) 09:58, 22 June 2022 (UTC)[reply]
    @Liuxinyu970226 Global ban is only put into effect after the consensus is formed. Being able to appeal and enforcing a ban are different things. Someone meets the requirement for proposing bans doesn't mean they will be banned. The current requirements are such that there is little focus on the nature of abuse. BRP ever 07:21, 23 June 2022 (UTC)[reply]
    @BRPever Have you ever read w:Armenia–Azerbaijan relations? The only reason for blocking Cekli829 is just based on this world's worst "diplomatic relationship". Liuxinyu970226 (talk) 03:48, 18 April 2023 (UTC)[reply]
    The issue here is that we are currently unable to even make a RfC for people who haven't been blocked on two projects. This has nothing to do with automatic global bans. People who have not been blocked on two projects, regardless of the amount of harm they have done to one project or the community as a whole, cannot have global ban requests written about them. This made sense back when we were only concerned with direct activity on Wikimedia projects, but now there are dozens of Wikimedia-related venues where harm can take place. Vermont 🐿️ (talk) 16:05, 24 June 2022 (UTC)[reply]
    @Vermont - is this list of venues published? I'd could be open to expanding it to include people that are also expelled from such places. — xaosflux Talk 09:34, 27 June 2022 (UTC)[reply]
    Not sure? We have affiliates (user groups, chapters, thematic orgs), mailing lists, phabricator, movement strategy forum, diff, various semi-official areas where off-wiki (but still wiki-related) harassment can take place. Causing harm to the movement to the level that warrants a global ban can really *happen* in a lot of places, but it takes quite a lot to pass that threshold. I'm not sure what good wording for this would be. Vermont 🐿️ (talk) 14:54, 27 June 2022 (UTC)[reply]
    @Vermont I'd think changing ...user is indefinitely blocked or banned on two or more projects venues ... should be an easy expansion, with a wiki-page of what constitutes those venues. — xaosflux Talk 15:03, 27 June 2022 (UTC)[reply]
    Indeed, we could give some examples, like: "venues (such as mailing lists, in-person events, online events, affiliate and community meetings, official and semi-official project communication channels, etc.)" depending on where the line should be drawn. Though...fundamentally, global policy has always aired on the side of "vague policy to give more room to variations in community consensus" over hyper-defined wording. This is one of the few holdouts. Vermont 🐿️ (talk) 15:10, 27 June 2022 (UTC)[reply]
    Not sure I'd want to include a "semi-official project communication channels" - so if I create #wikipedia-editathon-June2022 over on IRC and chanmode +b you, you are now "in scope". Basically, I only think places where a ban/block is enacted by a "community" (and/or their representatives) should be in scope. — xaosflux Talk 17:53, 27 June 2022 (UTC)[reply]
    Xaosflux, Sure, that works. Thoughts on a proposal for a wording change? Vermont 🐿️ (talk) 18:51, 27 June 2022 (UTC)[reply]
  • Support Support --Novak Watchmen (talk) 17:44, 23 June 2022 (UTC)[reply]
  • Oppose Oppose Many of us spend considerable time and energy dealing with harmful behaviour on various projects, and facilitating global ban nominations may seem like a good idea. However, in my experience, 99% of disputes can be resolved by engaging in a constructive dialogue, and usually a middle ground or compromise that is acceptable to everyone can be found with a bit of effort and open-mindedness. When a user is nominated for a ban, the goodwill to engage in a dialogue tends to erode very quickly, which in turn significantly lowers the chance of a diplomatic outcome of the dispute. I fear that making ban nominations easier will have an adverse, contra-productive effect. ArticCynda (talk) —Preceding undated comment added 19:31, 1 July 2022 (UTC).[reply]
    This allows global ban discussions to be started in cases where a user may not happen to be blocked on two wikis. This does not make ban nominations easier to pass, and bad requests would generally be quick-closed. Vermont 🐿️ (talk) 02:31, 28 August 2022 (UTC)[reply]
    @Vermont Eh, and why not do so? Sänger is just an example where they persis violating mw:CoC. Liuxinyu970226 (talk) 03:52, 18 April 2023 (UTC)[reply]
  • I would not be in favour of a general revocation of "at least 2 XXXXs", however, I think we could make a move from "2 local projects" to 2 Wikimedia spaces (projects, phab, usergroups, affiliates). I am a little reticent, as usergroups have had multiple issues with their conduct - blocking people on dubious grounds, and not having a suitable appellate system, involved issues etc. I'm open to discussion on that. Perhaps "Local projects, chapters/thematic organisations, phab, mediawiki", with a WMF event ban also counting as one of the two? Nosebagbear (talk) 23:08, 18 September 2022 (UTC)[reply]
  • Support Support. Someone who repetitively sabotages local de-syspopping by making false claim on Meta should probably be globally banned. If anyone wants any example of what I say, please email me privately.--Jusjih (talk) 01:21, 3 October 2022 (UTC)[reply]
  • Oppose Oppose - this requirement is neccesary to show that the user is both globally disruptive (otherwise no global steps are required) and unlikely to be redeemable (these bans are intended to be permanent). May be willing to reduce the "indefinitely blocked or banned", but proof of long-term disruption on multiple projects is an absolute. 93.172.254.44 08:17, 6 October 2022 (UTC)[reply]
  •  Weak support – I would support loosening the "blocked/banned in at least two projects" criterion, but for different reasons. I can think of a few limited circumstances where a user may need to be globally banned without having been sanctioned in two or more Wikimedia projects.
Consider, for example, an editor who has multiple contributions on several projects. It turns out that, on one specific project, most if not all of their edits consist of falsification of sources and fabricated content. This editor is rightly blocked and site-banned on that project only, but their contributions in other projects haven't been scrutinized. The logical step would be to globally ban this editor as a preventative measure, either because their other edits were similarly malicious or because their edits on that specific project might have already been translated on other wikis.
As another example, consider a user with cross-wiki contributions who engages in real-life harassment and makes deaths threats against other users in only one wiki. This user may technically not meet criterion #3 of the Global bans policy, but were such a ban take place, it would be SNOW-supported, and personally I would support bans of these users in principle.
In regard to "organized attempts to undermine the project's priorities" (like disinformation campaigns), I think that these should be treated similarly like sockpuppetry and other forms of disruptive editing. If a user is attempting to promote disinformation on a specific wiki, then of course this user should be blocked and banned there, but as long as they aren't being disruptive on other projects I don't see the need to implement a global ban.
Regards, Dsuke1998AEOS (talk) 15:13, 11 December 2022 (UTC)[reply]
@Dsuke1998AEOS Then how do you resolve Sänger's issues? Liuxinyu970226 (talk) 07:12, 2 August 2023 (UTC)[reply]
  • Support Support Reasonable proposition. fr33kman 17:13, 27 October 2023 (UTC)[reply]
  • Strong oppose From what I've seen, global ban RFCs usually amount to a plebiscite rather than a serious, structured effort to determine whether the user in question is eligible and violated site policy in such a way that a global ban is necessary. Even these reasonable constraints are interpreted loosely or ignored altogether. I see no reason to make it even easier. Perhaps it's relevant also that meta does not generally override other projects in the opposite way, that is, by unblocking users or arbitrating appeals when the blocking sysops/admins have acted contrary to wikimedia policy. Users are usually told to use the local appeals process wherever they were blocked, and that meta cannot get involved, or something to that effect. AP295 (talk) 17:04, 7 February 2024 (UTC)[reply]
  • Oppose Oppose As long as requests for global bans tend toward being mere numerical votes rather than requests for comment proper (can someone refute this?), formal hurdles for starting such a numerical vote are very valuable. A problem that only concerns one wiki is soluble via a measure pertaining to that wiki only, even if the solution were to come from Meta; a global ban is then not appropriate and is excessive. On a specific case: a capture of, say, Croatian wiki does not require a global ban. Other solutions are available. One solution is to ban indefinite blocks on a project suspected to be captured, limiting the block length to, say, one year. And that solution would be much more effective since it would impact behavior of all admins, not just selected one. Another solution is externally-forced desysop of Kubura (if this is what we are talking about): he would be able to contribute, but not harass others with powertools. --Dan Polansky (talk) 11:43, 8 February 2024 (UTC)[reply]
  • Support Support - I see no reason why this procedural issue should get in the way. Theknightwho (talk) 20:51, 8 February 2024 (UTC)[reply]
  • Oppose Oppose Agree entirely with 93.172.254.44 * Pppery * it has begun 03:01, 9 February 2024 (UTC)[reply]
Retrieved from "https://meta.wikimedia.org/w/index.php?title=Requests_for_comment/Global_ban_requirements&oldid=26198169"