Wikipedia:Edit filter/Requested
This page can be used to request edit filters, or changes to existing filters. Edit filters are primarily used to address common patterns of harmful editing.
Private filters should not be discussed in detail. If you wish to discuss creating an LTA filter, or changing an existing one, please instead email details to wikipedia-en-editfilters
lists.wikimedia.org.
Otherwise, please add a new section at the bottom using the following format:
== Brief description of filter ==
* '''Task''': What is the filter supposed to do? To what pages and editors does it apply?
* '''Reason''': Why is the filter needed?
* '''Diffs''': Diffs of sample edits/cases. If the diffs are revdelled, consider emailing their contents to the mailing list.
~~~~
Please note the following:
- Edit filters are used primarily to prevent abuse. Contributors are not expected to have read all 200+ policies, guidelines and style pages before editing. Trivial formatting mistakes and edits that at first glance look fine but go against some obscure style guideline or arbitration ruling are not suitable candidates for an edit filter.
- Filters are applied to all edits on all pages. Problematic changes that apply to a single page are likely not suitable for an edit filter. Page protection may be more appropriate in such cases.
- Non-essential tasks or those that require access to complex criteria, especially information that the filter does not have access to, may be more appropriate for a bot task or external software.
- To prevent the creation of pages with certain names, the title blacklist is usually a better way to handle the problem - see MediaWiki talk:Titleblacklist for details.
- To prevent the addition of problematic external links, please make your request at the spam blacklist.
- To prevent the registration of accounts with certain names, please make your request at the global title blacklist.
- To prevent the registration of accounts with certain email addresses, please make your request at the email blacklist.
Overwriting of birth range templates
[edit]| This section is pinned and will not be automatically archived. |
- Task: Warn or disallow when an edit replaces a {{birth based on age as of date}} template with a date or year, without simultaneously adding a URL or reference tag
- Reason: This "based on age" template is a strong indicator that Wikipedia editors have already tried and failed to find an exact, reliably-sourced birth date for a person, and have only been able to confirm a two-year date range. Readers sometimes step in to replace these with a particular year or date, offering no source, often just asking IMDb or Google. An unsourced DOB is obviously against WP:BLPDOB, and a "based on age" template suggests that any added date (even if correct) has not been
widely published by reliable sources
. - Diffs: eg. [1], [2]. Belbury (talk) 09:12, 17 March 2026 (UTC)
New users commenting out blocks of text
[edit]| This section is pinned and will not be automatically archived. |
- Task: If new users add hidden comments to articles (using <-- --> plus !, I can't figure out how to escape it) that encompass large blocks of text, it should be flagged.
- Reason: This is a sneaky way of hiding text without appearing as a removal of content and there are very few valid reasons to do this.
- Diffs: [3] SmartSE (talk) 23:08, 18 March 2026 (UTC)
- I'm looking into implementing this. Daniel Quinlan (talk) 00:07, 19 March 2026 (UTC)
Log (or warn) assessing articles as A-class
[edit]- Task: Log when the class parameter in a WikiProject template is changed to
A. Perhaps also give a warning that A class should not be added unilaterally (and the highest class that can be added is B). - Reason: Per Wikipedia:Content assessment/A-Class criteria, A class should not be added unilaterally. This is not clear to editors who haven't read the relevant pages and see A as a continuation of the stubstartCB system.
- Diffs:
- Special:Diff/1357446590
- Special:Diff/1318092734
- Special:Diff/1324394406
- Special:Diff/1321693846
- Special:Diff/1353386910
- and that's just from looking through one category
Times like these I wish I was still an EFM... charlotte 👸♥ 02:29, 3 June 2026 (UTC)
- @Queen of Hearts: I could kind of see this for non-extended confirmed, though it doesn't seem practical to apply this to every single editor. Do you have any ideas for limiting this down perhaps? EggRoll97 (talk) 18:06, 30 June 2026 (UTC)
Removal of AI tags
[edit]- Task: Log instances where an edit has removed {{AI-generated}} (another filter for {{Db-g15}} and {{Prod llm}} would be desirable as well)
- Reason: Sometimes an editor will quietly remove an AI-generated tag without addressing any of the issues, and it is challenging to keep track of these instances. This was briefly discussed here
- Diffs: [4] [5], can hopefully provide some diffs re deletion tags shortly
Kowal2701 (talk, contribs) 16:12, 7 June 2026 (UTC)
- For a recent example of improper LLM deprodding that such filters could help monitor, see WP:AINB#~2026-33518-93. fifteen thousand two hundred twenty four (talk) 22:30, 7 June 2026 (UTC)
- More improper deprod examples that I only caught because one pinged my watchlist [6][7][8][9], the subsequent conversation on my talk page made it clear the editor was treating them as normal prods [10]. fifteen thousand two hundred twenty four (talk) 08:20, 10 June 2026 (UTC)
- Another,
All the links I opened were working.
(following added 01:24, 14 June 2026 (UTC)) Now sockblocked [11]. fifteen thousand two hundred twenty four (talk) 06:26, 13 June 2026 (UTC)- another: "I asked ChatGPT whether this was AI and they said it could be either" Gnomingstuff (talk) 21:58, 28 June 2026 (UTC)
- Another,
- More improper deprod examples that I only caught because one pinged my watchlist [6][7][8][9], the subsequent conversation on my talk page made it clear the editor was treating them as normal prods [10]. fifteen thousand two hundred twenty four (talk) 08:20, 10 June 2026 (UTC)
- In both examples, there were clear edit summaries regarding removing the tag, and both removals were made by established editors. Would this suggestion address the problems with AI articles, or would it just provide a convenient list of battlegrounds for existing debates regarding LLMs? For the latter, I recommend: Wikipedia talk:Writing articles with large language models Drew Stanley (talk) 23:34, 8 June 2026 (UTC)
- My bad, honestly I got the diffs from someone else and didn't look at them, they said those were ones they remembered most, presumably because it caused disputes. That's pretty rare though. Here's some diffs for what I mean re the above (some are on now-deleted articles, I got these from the archives at WP:AINB): [12] [13] [14] [15] [16] [17] [18], Special:Contributions/Tired Carrot (WP:AINB#User:Tired Carrot) is what spurred me to finally make this request Kowal2701 (talk, contribs) 00:05, 9 June 2026 (UTC)
- A more conservative option would be AI tags removed without edit summaries and/or by users with few edits. I think with this filter, conservative is better, there are a lot of valid reasons to remove AI tags (cleanup being done, no major edits since before 2023, the person who added the tag changing their mind, etc.) Gnomingstuff (talk) 17:53, 23 June 2026 (UTC)
- I believe that @Kowal2701 got those diff links from me. The tag should not have been removed in either of those case, they're both examples of incorrect removals of the tag. And regardless of that, they show what changes a regex for a filter like this needs to match, the motivation for removing them doesn't factor into the regex. ‑‑gurkubondinn 19:22, 23 June 2026 (UTC)
- While I am open to Gnomingstuff's suggestion, I disagree that these are "quiet" removals as they are clearly acknowledged in the edit summaries, and I maintain that this feels battleground motivated. Most of the new diffs given by Kowal have been deleted/are inaccessible. Drew Stanley (talk) 16:51, 1 July 2026 (UTC)
- ... maybe you've heard of WP:AGF? Kowal2701 (talk, contribs) 16:53, 1 July 2026 (UTC)
- it isn't bad faith to ask for accessible evidence for a proposal that you are pushing for with "ASAP" levels of urgency Drew Stanley (talk) 16:55, 1 July 2026 (UTC)
- Wikipedia:Assume the assumption of good faith Drew Stanley (talk) 16:59, 1 July 2026 (UTC)
- There have been plenty of accessible diffs given here, and admins and most EFMs have access to the deleted ones. I get why the first report may have given you this impression, but not why the following hasn't dispelled it. Also see [19] Kowal2701 (talk, contribs) 17:03, 1 July 2026 (UTC)
- Wikipedia:Assume the assumption of good faith Drew Stanley (talk) 16:59, 1 July 2026 (UTC)
- it isn't bad faith to ask for accessible evidence for a proposal that you are pushing for with "ASAP" levels of urgency Drew Stanley (talk) 16:55, 1 July 2026 (UTC)
- Im personally more concerned with improper deprodding, as there's a specific process that should be followed for that, but there are tons of improper ai tag removals as well, just look through WP:AI noticeboard. I fail to see how this is battleground motivated. For what POV? InfernoHues (talk) 17:17, 1 July 2026 (UTC)
- Tbh I think it's easy to perceive people at AINB as an anti-AI faction, it's something I've been a bit concerned about actually developing, there's plans to merge the Discord servers for WP:AIC and WP:AIT so it's more pluralistic, I'm not sure what else can be done other than to be conscious of it Kowal2701 (talk, contribs) 17:21, 1 July 2026 (UTC)
- I think people would be less likely to have that impression if they went to AINB, picked a random article and checked its source-text integrity. Agree that not much can be done however. InfernoHues (talk) 17:27, 1 July 2026 (UTC)
- Tbh I think it's easy to perceive people at AINB as an anti-AI faction, it's something I've been a bit concerned about actually developing, there's plans to merge the Discord servers for WP:AIC and WP:AIT so it's more pluralistic, I'm not sure what else can be done other than to be conscious of it Kowal2701 (talk, contribs) 17:21, 1 July 2026 (UTC)
- ... maybe you've heard of WP:AGF? Kowal2701 (talk, contribs) 16:53, 1 July 2026 (UTC)
- While I am open to Gnomingstuff's suggestion, I disagree that these are "quiet" removals as they are clearly acknowledged in the edit summaries, and I maintain that this feels battleground motivated. Most of the new diffs given by Kowal have been deleted/are inaccessible. Drew Stanley (talk) 16:51, 1 July 2026 (UTC)
- I've made a filter to try to catch removal of {{AI-generated}}. The optional line can be added to restrict the filter to only edits without summaries, though I'd prefer logging them all. If this is well liked I can write the regex for deLLMPRODing as well.
Filter
|
|---|
equals_to_any(page_namespace, 0) &
(
action == "edit" &
summary = "" & //optional
removed_lines rlike '\{\{\s*[Aa][Ii][-_ ]generated(?:\s*\||\s*\}\})' &
) |
- I would definitely support having something like this for AI cleanup patrol reasons. I've noticed many times, when AI was used to promote a person/company, a UPE will do some minor changes, remove the tag, and get checkuser blocked. See the history of HAUI as an example. InfernoHues (talk) 18:48, 23 June 2026 (UTC)
- I'd like to see this filter be set to 'warn' instead of just 'log'. Specifically, it should give a simple, educational message that says something like "Since March 2026, the use of chatbots, AI, or LLMs to generate or rewrite article content is prohibited for most purposes. See WP:NOLLM for more information. Please do not remove the maintenance template until the problem has been resolved." WhatamIdoing (talk) 20:26, 23 June 2026 (UTC)
- one of the problems is that a lot of people think that "i cant see any ai here" is the same as "the problem has been resolved". or other bizarre reasons like "this has been reviewed by npp" or "i clicked all of the links and they worked for me". but i agree with you still, i think the edit filter should be set to "warn" (after letting it run in "log" mode for a while to watch for false positives first). ‑‑gurkubondinn 20:33, 23 June 2026 (UTC)
- I don't really like the idea of warning people who have a high chance of not having done anything wrong (legitimate tag removals). If logging goes well, we can probably change to tagging, so that removals are more visible. InfernoHues (talk) 20:38, 23 June 2026 (UTC)
- A technical warning step doesn't imply that you've done something wrong, it's just a step to confirm that you intended to do what you just did. For example how the edit filters for the WP:AWARE templates used to work, you got a warning reiterating that you had to check if the user had previously been notified or not (this filter has now been removed or disabled). ‑‑gurkubondinn 11:39, 1 July 2026 (UTC)
- I don't really like the idea of warning people who have a high chance of not having done anything wrong (legitimate tag removals). If logging goes well, we can probably change to tagging, so that removals are more visible. InfernoHues (talk) 20:38, 23 June 2026 (UTC)
- one of the problems is that a lot of people think that "i cant see any ai here" is the same as "the problem has been resolved". or other bizarre reasons like "this has been reviewed by npp" or "i clicked all of the links and they worked for me". but i agree with you still, i think the edit filter should be set to "warn" (after letting it run in "log" mode for a while to watch for false positives first). ‑‑gurkubondinn 20:33, 23 June 2026 (UTC)
- I'd like to see this filter be set to 'warn' instead of just 'log'. Specifically, it should give a simple, educational message that says something like "Since March 2026, the use of chatbots, AI, or LLMs to generate or rewrite article content is prohibited for most purposes. See WP:NOLLM for more information. Please do not remove the maintenance template until the problem has been resolved." WhatamIdoing (talk) 20:26, 23 June 2026 (UTC)
- Sorry but is it possible to do this ASAP? I've just done like 250 PRODs and it's a massive pain to have to keep checking whether they've been improperly removed like [20]. I'd like to keep working through the backlog but can't really without this Kowal2701 (talk, contribs) 11:35, 1 July 2026 (UTC)
- Two more recent ones [21][22]. It would be nice if I didn't need to add hundreds of tagged pages to my watchlist. fifteen thousand two hundred twenty four (talk) 13:54, 1 July 2026 (UTC)
- We could make it a WP:STICKYPROD. WhatamIdoing (talk) 23:13, 1 July 2026 (UTC)
- Are you running a log, like the CSD logs? That makes it pretty easy to see how many/which ones are being kept. WhatamIdoing (talk) 23:14, 1 July 2026 (UTC)
- Two more recent ones [21][22]. It would be nice if I didn't need to add hundreds of tagged pages to my watchlist. fifteen thousand two hundred twenty four (talk) 13:54, 1 July 2026 (UTC)
Monitoring at 1407 – PharyngealImplosive7 (talk) 14:56, 1 July 2026 (UTC)
- Thank you, would it be possible to separate these into different filters or indicate which of {{ai-generated}} or {{prod llm}} was removed? Both deal with LLM use but have very different requirements for application and removal. ai-generated is like a typical maintenance tag, while prod llm has it's own guideline with specific enforceable requirements that must be met before it can be removed. fifteen thousand two hundred twenty four (talk) 15:02, 1 July 2026 (UTC)
- +1 InfernoHues (talk) 15:30, 1 July 2026 (UTC)
Done See 1407 (hist · log) and 1408 (hist · log) – PharyngealImplosive7 (talk) 19:50, 1 July 2026 (UTC)
- Thanks again, and sorry to add another thing, but sysops, patrollers and EC editors shouldn't be excluded, especially not for 1408. [23][24][25][26][27] are all improper deprods that meet one of the exclusion criteria. Banner blindness can happen to any of us, and many editors see "prod" and assume it works like normal WP:PRODs.
- 1407 may have too much noise otherwise, but I think it makes sense to start with an unfiltered feed there too so we can gauge things (and I've definitely seen autopatrolled users improperly remove ai-generated tags, such as in [28] which is linked above). fifteen thousand two hundred twenty four (talk) 05:17, 2 July 2026 (UTC)
- Neither of the filters should be excluding XC, patrollers or sysops. Could this be fixed please? ‑‑gurkubondinn 14:21, 2 July 2026 (UTC)
- Agree with the above, at the very least EC shouldn't be excluded, it's really easy to rack up edits if you use LLMs. InfernoHues (talk) 14:31, 2 July 2026 (UTC)
- Done as well – PharyngealImplosive7 (talk) 14:39, 2 July 2026 (UTC)
- 1408 failed to log a test removal and actual removal (a proper one). My guess is since
added_linesis a string array the matching isn't multiline? If that's the issue then\{\{\s*prod[-_ ]llm\/dated\s*[|\n]should be sufficient (I'm assuming added_lines retains ending newlines, but have no way to test), it wouldn't catch if the closing }} is removed, but should catch the template target being changed by inserting text after it like{{prod llm/dated off. fifteen thousand two hundred twenty four (talk) 05:31, 3 July 2026 (UTC)- I've fixed the newline issue using your suggested regex – PharyngealImplosive7 (talk) 18:46, 3 July 2026 (UTC)
- Looks good, thanks for adding a BLAR exception too. fifteen thousand two hundred twenty four (talk) 04:06, 4 July 2026 (UTC)
- @PharyngealImplosive7 just wondering, would it be worth making the "redirect" text in line 10 case-insensitive, so that either #REDIRECT or #redirect is exempted, like this:– Epicgenius (talk) 16:06, 6 July 2026 (UTC)
added_lines rlike "\#(REDIRECT|redirect) \[\[.+\]\]" &
- Done. I used
irlikeinstead ofrlike– PharyngealImplosive7 (talk) 16:59, 6 July 2026 (UTC)- Thanks, much appreciated. Accessedgrant (Epicgenius mobile alt) (talk) 12:33, 8 July 2026 (UTC)
- @PharyngealImplosive7 (sorry to ping you again), is it possible to include removal of tags in draftspace? Often AI-generated articles get draftified as a quick fix, and sometimes people move them back to mainspace later without doing a proper review/rewrite. I've started systematically tagging drafts, but it'd help if they were included here. Maybe userspace as well because we encourage people to userfy articles Kowal2701 (talk, contribs) 12:44, 19 July 2026 (UTC)
- Done. I used
- @PharyngealImplosive7 just wondering, would it be worth making the "redirect" text in line 10 case-insensitive, so that either #REDIRECT or #redirect is exempted, like this:
- Looks good, thanks for adding a BLAR exception too. fifteen thousand two hundred twenty four (talk) 04:06, 4 July 2026 (UTC)
- I've fixed the newline issue using your suggested regex – PharyngealImplosive7 (talk) 18:46, 3 July 2026 (UTC)
- 1408 failed to log a test removal and actual removal (a proper one). My guess is since
- Done as well – PharyngealImplosive7 (talk) 14:39, 2 July 2026 (UTC)
- +1 InfernoHues (talk) 15:30, 1 July 2026 (UTC)
- Thank you, would it be possible to separate these into different filters or indicate which of {{ai-generated}} or {{prod llm}} was removed? Both deal with LLM use but have very different requirements for application and removal. ai-generated is like a typical maintenance tag, while prod llm has it's own guideline with specific enforceable requirements that must be met before it can be removed. fifteen thousand two hundred twenty four (talk) 15:02, 1 July 2026 (UTC)
Possible caste-related edits by non-extended-confirmed users
[edit]- Task: Tag all edits that add (or use edit summaries containing) words related to caste in South Asia. It would apply to edits by non-extended-confirmed users on mainspace pages.
- Reason: WP:CT/CASTE places all caste-related topics under the extended-confirmed restriction. WP:CT/CASTE violations are often not caught, and they're often WP:BLP violations as well.
- Diffs: [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50]
- The following case-insensitive regex has some basic/general caste terms. I'm still working on a regex that includes a bunch of more specific castes, but I think this would be a good start and would catch a high proportion of caste-related edits. I've tried to be relatively conservative and exclude anything that's common as a surname or that's likely to get a lot of false positives for other reasons.
\b((caste|subcaste|jati|jāti|brahmin|bramhin|bhusura|brahimin|bramin|kshatriya|kashtriya|khasatriya|khsatriya|ksatriya|kṣatriya|kshartriya|kshathriya|kshatria|kshattriya|vaishya|vaiśya|vaisya|vaishiya|vyshya|shudra|śūdra|dalit|harijan|harizan|obc|pasmanda|ajlaf|jat|rajput)s?|brahmans|backwards? class(es)?|untouchability)\b
Cadddr (talk) 17:44, 8 June 2026 (UTC)
- +1 with above. Zalaraz (talk) 18:07, 8 June 2026 (UTC)
- Support This is useful per Cadddr. If you can, alphabetize the lists For the longer list: chenchu (see Kannappa) and choda, chudasama (see Kathi people). This list could also include gazetteers and colonialists like James Tod which are used, usually by editors without EC, for caste-related edits (reasoning here). Drew Stanley (talk) 23:29, 8 June 2026 (UTC)
- I've also seen "antal" Drew Stanley (talk) 03:41, 9 June 2026 (UTC)
- This seems like it might be a good tagging filter, but I think we're going to need a much longer list of castes and a list of common sources used for these edits, especially castes or sources that are commonly used in problematic edits or by LTAs. More example diffs would also be very helpful. Note that some expressions like "obc" might not be usable in a filter due to false positives, but perhaps those might be usable in a more complex expression looking for specific phrases. Daniel Quinlan (talk) 20:37, 9 June 2026 (UTC)
- Yes, I'm still working here on making a longer list of castes, though it'll be a little while before it's ready. My current process is to search Wikipedia for the term to see if there's a significant number of false positives. For ones with false positives, I'm adding it to a more complex regex, which is currently of the form
\b(CASTE LIST HERE...)(s\b|[\s-]+(community|communities|tribe|person|people|class|cast|population|clan|tribe|man|woman|men|women|group|origin|descent|reservation|quota)). - The reason I started by proposing the short, general list is that to my surprise, the vast majority of the diffs I found would be caught by the short list. But it's been a little harder than I expected to find diffs, so that might just be a property of the diffs I've found so far. Is there an easy way to just search through all recent edit summaries for phrases like "WP:CT/CASTE", "WP:CT/SASG", or "WP:ECRCASTE"?
- I included "OBC" initially because when I searched for it, there were fewer false positives than I expected. Though as I'm looking at it again, maybe it's still too many. It looks like the plural ("obcs") has basically no false positives. Cadddr (talk) 21:35, 9 June 2026 (UTC)
- And yeah, the suggestion by you and Drew Stanley to include sources too seems like a good one. I don't know much about that angle, though. Cadddr (talk) 21:41, 9 June 2026 (UTC)
- Yes, I'm still working here on making a longer list of castes, though it'll be a little while before it's ready. My current process is to search Wikipedia for the term to see if there's a significant number of false positives. For ones with false positives, I'm adding it to a more complex regex, which is currently of the form
- The only term here I'm not sure about is "harizan" – @Cadddr, could you please explain that one? Is it a common misspelling of "harijan"? The rest seem appropriate to include, though I think Drew is correct that the list could be much longer. Toadspike [Talk] 20:48, 9 June 2026 (UTC)
- @Toadspike I threw in "harizan" because I saw that misspelling being used somewhere. (And I figured it makes sense theoretically for "z" and "j" to get mixed up, since Urdu speakers distinguish those two phonemes but Hindi speakers generally don't.) I'm not sure how common the misspelling is. Cadddr (talk) 21:39, 9 June 2026 (UTC)
- We could condense the suggested regex to:
\b((sub)?caste|j[aā]ti|bra(hm|mh?)[ia]n|ka?(sh|ṣ|s)a?r?tt?riya|vai(sh|ṣ|s)ya|(ś|s|sh)[ūu]dra|dalit|harijan|ajlaf|jat|rajput)s?. I also searched up "harizan" on wikipedia and it doesn't seem to refer to the caste and seems to be a name, so I haven't included it. Others are right that this is a pretty small limited set of words and that it should be expanded – PharyngealImplosive7 (talk) 18:56, 10 June 2026 (UTC)- One small thing: it was intentional that I included "Brahmans" plural (since that pretty much has to refer to Brahmins) but not the singular "Brahman", which is ambiguous and more often refers to the Hindu metaphysical concept of Brahman (see Brahman § Comparison of Brahma, Brahman, Brahmin and Brahmanas). Cadddr (talk) 19:42, 10 June 2026 (UTC)
- I think then we could change
bra(hm|mh?)[ia]ntobra(hm|mh?)inand just addbrahmansat the end. – PharyngealImplosive7 (talk) 19:46, 10 June 2026 (UTC)- Agreed. Cadddr (talk) 19:47, 10 June 2026 (UTC)
- I think then we could change
- One small thing: it was intentional that I included "Brahmans" plural (since that pretty much has to refer to Brahmins) but not the singular "Brahman", which is ambiguous and more often refers to the Hindu metaphysical concept of Brahman (see Brahman § Comparison of Brahma, Brahman, Brahmin and Brahmanas). Cadddr (talk) 19:42, 10 June 2026 (UTC)
- I'll let you all know when I've finished making the larger regex at User:Cadddr/sandbox/Caste edit filter#More specific castes. If anyone wants to help, please do! Cadddr (talk) 19:45, 10 June 2026 (UTC)
A longer list and more diffs
[edit]Alright, I'm finally done creating a much larger version of the regex. For ease of reading, I've split it into 5 parts, which should be joined together, separated with pipes (|):
- terms that don't need context:
(?i)\b(24[\s\-]*manai|(24|twenty[\s\-]*four)[\s\-]*m[\s\-.]*t[\s\-.]*c[\s\-.]*|96[\s\-]*(kuli|maratha)|[aā][nñ]ja[nṇ][aā][\s\-]*chaudhari|adi[\s\-]*andhra|adi?[\s\-]*(dharmi|dravida|karnataka)|agamudayar|ahinda|ajlaf|akkasali|akkasaliga|ambalavasi|ambattar|anamuk|āñjaṇā|ar(ay|wa)[\s\-]*mala|(arun[aā]ttu|chozhia|devendrakula|kaa?rkaa?r?thaa?r|kongu|mottai|thuluva|sri[\s\-]*lankan)[\s\-]*vell?[āa]la[nr]?|arundhatiyar?|arunthathiyar|asadaru|(ati[\s\-]*)?(ś|s|sh)[ūu]dra|ayya(na)?[vw]ar|bagh[eu]ti|bahun|baj(aniya|gi)|balagai|ban(ajig|chad)a|banbansi|bansphor|barujibi|(besati|chatradhari|sapure)[\s\-]*mal|bhambi[\s\-]*(khalpa|rohit)|bhargava[\s\-]*dakaut|bhumihar|bora[\s\-]*karki|brahmakshatriya|bh?ra(hm|mh?)in|budabukkala|budubudiki|chakkiliya[nr]|chal[au]?v[aā]di|(chandraseniya([\s\-]*kayastha)?|pathare|kanchole|davane)[\s\-]*prabhu|chattadi[\s\-]*sathtavar|cheetah[\s\-]*mehrat|chenna[\s\-]*dasar|chhapri(band|basi|wale)|chhimb[ae]|chindollu|(chitraguptavanshi|north[\s\-]*indian|bengali)[\s\-]*kayastha|chuhra|chutor[\s\-]*mistri|creamy[\s\-]*layer|c[ūu]nv[āa][ḷl]iy[āa]|dagoli|dahait|dalit|dandachhatra[\s\-]*majhi|dangashia|dankha?ni[\s\-]*mang|dasia[\s\-]*bauri|[ḍd]h[ēe][ḍd]h|dedh[\s\-]*(barot|sadhu)|deshastha|desua[\s\-]*pana|devendrakulathan|dhankia|dhaogri|dharikar|dharival|dhegu[\s\-]*megu|dhenuar|dhuai|dh[ūu][ḷl]adh[ōo]y[āa]|dhyar|dokkalwar|doomna|dosadha?|duria[\s\-]*dom|ellamala?[vw]ar|ezhava|gadaria|gandhabanik|gandhila|gandil[\s\-]*gondola|(gangadikara|morasu|kunchitiga|namdhari|tulu|kodagu|nonaba|sadar)[\s\-]*[vw]?[ao]kkaliga[rn]?|ganti[\s\-]*chores|garancha|garmatang|general[\s\-]*category[\s\-]*candidate|gh[āa][ñn̄n]c[īi]|ghamaila|ghantarghada|ghogia|ghusuria|giarah|godag(ali|ula)|gorait|halalkhor|halasvar|halenadu[\s\-]*karnataka|halleer|handi[\s\-]*jogi|harijan|helavaru|hol[ae]ya|hulasvar|hurkiya|jaangid|jaggali|jalia[\s\-]*kaibartt?a|jalkeot|jambuvulu|jamphoda|(jatav|jatia|pardeshi)[\s\-]*chamar|jhalo[\s\-]*malo|kaarkaathaar|kabbaliga|kabirpanthi[\s\-]*j[ou]ll?ah[ae]|kachhandha|kadaiyan|kada?[vw]a[\s\-]*patidar|kalabaz|kalari[\s\-]*(kurup|panicker)|kamati[\s\-]*mochi|kandabari|kanugh|kapadia[\s\-]*sansi|kapariya|karoack|ka?(sh|ṣ|s)a?r?tt?riya|ka[ūu]m[ūu]ti|kepmari|keyot|khairaha|khandayat|khav[āa]sas|(kodava|jamma)[\s\-]*maaple|koiri|kolupula?vandlu|kondayankottai|kooch[\s\-]*band|koppula[\s\-]*velama|korachar|korait|koravar|kovil[\s\-]*padagar|kovil[\s\-]*thurayar|kubutar|kuchband(hia)?|kudmi[\s\-]*mahato|kunjra|kurariar|kuruba|kuthaliya[\s\-]*bora|lala[\s\-]*baig|lal[\s\-]*begi|leva[\s\-]*patel|l[īi]maciy[āa][\s\-]*valand|lingader|ling[āa]yath?|madar[’']i|mada[rs]i[\s\-]*kuruva|madiga[\s\-]*dasu|mahayavanshi|mahishya|majhabi|majhbi|mala[\s\-]*(ayawaru|dasari|dasu|hannai|jangam|masti|sale|sanyasi)|mal[\s\-]*muslim|mallaah|manai[\s\-]*telugumang[\s\-]*(gar[ou]di|mahash?i|garudi)|mapilla|maratha[\s\-]*clan|maru[\s\-]*vankar|matha[\s\-]*pulayan|matia[\s\-]*kela|(mazhabi|rai|rangreta)[\s\-]*sikh|meghval|mehtor|menghvar|minimadig|mitha[\s\-]*ayyalvar|mochigar|mogeyar|mukkulathor|mukkuvar|mundapotta|musahar|nabhadassia|nadar[\s\-]*climber|nagarathar|nalakeyava|nalkadaya|nalua[\s\-]*kela|namasudra|naribut|nashya[\s\-]*(sekh|shaikh)|navdigar|padannan|palluvan|pambada|panchamsali|panchkalshi|panikhia|pannayeri[\s\-]*nadar|panniandi|pantanti|parahiya|parai?yar|patratanti|phe?rera|pulaya[\s\-]*vettuvan|puthirai[\s\-]*vannan|r[aā]jopadhy[aā]y[aā]|radhe[\s\-]*mang|rajuad|rajuala|rajuar|ramdasia|raneyar|ranigar|relli[\s\-]*chachandi|rujjhar|s[āa]mantan|sabakhia|sabhaiyar|(sadhu|saluppa|elur)[\s\-]*chetty|sakkiliyar|sanaurhiya|sanhal|sansiya|sapua[\s\-]*kela|sarasvata|saraswat|sarbhangi|sarehde|saryara|satanaiah|satani[\s\-]*vaishnavulu|sathata[\s\-]*aiyyar|sathvara|sattada[\s\-]*ekaki|sattadavar[\s\-]*purohitar|satvik[\s\-]*vaishnava|sat[vw]ara|sauntia|setty[\s\-]*balija|sevayat|shahannava[\s\-]*(kuli|maratha)|shershabadia|sidhanar|sidhria|sikligar|silawat|sillekyatha|sindhi[\s\-]*mochi|sindhollu|singiwala|sirkiband|siryare|(sub)?caste(is[mt])?|surjyabanshi|surjyaramnami|swakula[\s\-]*sali|tagadhari|tamudia|tantubaya|tarmali|telugu[\s\-]*chetty|telugu[\s\-]*mochi|(thakor|thakarda)[\s\-]*koli|thathagar|thummagunta[\s\-]*drāvida|tirbanda|(tulu|arebhashe)[\s\-]*g(ow|au)da|turaiha|turi[\s\-]*barot|v[āa](ṇ|nn?)iy[aā]r?|vai(sh|ṣ|s)ya|vaḷā|valamiki|valanjiyar|vanjiri|vataliy?a[\s\-]*prajapati|vathiriyan|venkatapura[\s\-]*vaishnav[āa]|vettuva[\s\-]*gounder|vishnu[\s\-]*archaka|voddar|wayanad(an)?[\s\-]*pulayan|yaduvanshi|yellammalawandlu|zadmalli|zamral)s?\b
- terms that need either pluralization or additional context words afterwards (usually because they're also surnames):
\b([aā]h[iī]r(war)?|agariya|agg?(ra|[ae]r)wall?a?|aheer|aheri|ahluwalia|(a|i|ai)y[ae]ngar|ajila|arain|arora|asati|avarna|b[aā]liga|b[aā]ro[ṭt]|ba(bar|hel)iy?a|ba(dh|[dr]a)ik|bagal|baghban|bai(ga|rwa|swar)|bal(ija|miki?)|bananju|banjara|ban?sor|barahsaini|batwal|bawaria|bh[aā]vas[aā]r|bh?ambhi|bhambi|bh(aa?|ā)nd|bhandari|bhang[iī]|bhanushali|bharwad|bhatti|bhishti|bisati|brahman|chamaa?r|cham(bh|g)?ar|chandal(am?)?|chandel|ch?[aā]ra[nṇ]|charmar?kar|chattada|chenva|cher[au]ma[nr]?|chetti|chettiar|chhetri|chhimpa|chhipi|chikwa|dabgar|daha(ya)?t|dakkal(iga)?|dand[aā]si|dara?j[īi]|dhanak|dhangar|dhangri|dhanuk|dhanwar|dharhi|dharkar|dhed|dhibara|dhiman|dhob[aīi]|dhogri|dhor|dh?usadh|dhusia|dohor|dombari|[ei]diga|ezhuthachan|g[āa]n(iga|dla)|gadariy?a|garo[ḍd][āa]|garodi|gavari?a|gh[āa]siy?a?|ghantra|gharr?ami|ghosi|go[áa]l[áa]|god[āa]ri|godhi|gonrhi|g[ōo]sangi|gopal|goswami|gounder|gowari|gujjar|gurav|gurjar|gwalvanshi|hajjam|halwai|hatkar|helava|hol[aei]ya|iyer|j(a|aa|ā)ti|jaa?tt?|jangid|jatava?|jatia|jhal[ao]|(jik|lah)ar|joshi|j[ou]ll?ah[ae]|jyapu|k[āa]lin[gj]i|kachhi|kada?[vw]a|ka[ḍd]iy?[āa]|kah[aâ]r|kaibartt?a|kaikadi|kaikala|kakkalan|kakkan|kakkayya|kalita|kal[lw]ar|kamad|kamma|kammalar|kamoh|kanakkan|kanbi|kaniyar|kanjar|kankayya|kans[āa]r[āa]|kaora|kapali|karan|kathi|kathodi|kavara|kayastha|keot|keuta|kewat|kh[āa]car|khadala|khadit|khaira|khandait|khandelwal|khangar|kharol|khat(ee|i)k|khatri|khodalo|khorot|khosar|khotta|kirar|kodalo|kodava|k(o[ei]|eo)ri|kolh?i(yan)?|komarpanth?|koodan|k[ōo][ḷl][īi]|k[ōo]ma[ṭt]i|koosa|kootan|korachar?|korar|kotegar|k[ou]rava[nr]?|kudumban|kudum?bi|kulala|kumb?h[āâa]r|ku(mb|mbh|mh)[āa]ra?|kummari|kunbi|kurumbi|kurunga|kushwaha|k[ūu]dan|k[ūu]ma[ṭt]i|lababb(een|ai|ei)|labab(ee|[iī])n|l(ab|av|oh|ubh|uv)ana|labbay|laheri|lakhera|lebb(ai|ay)|lodh[ai]?|lohar|lonari|lubb(ay|ee|ye)|luh[āa]r|m[aā]diga?|machigar|mahar|mahasha|mahatam|mahesh[vw]ari|mahisyada|mahuria|mahyavansh?i|majhwar|malkana|mallah|manihar|mans(oo|u)ri|mappill?a|marava[nr]?|marecha|mashteen|mavilan|mazhabi|megh|meghwa[lr]|meht?ar|mirdha|mogaveera|moghia|motisar|muchi|mudaliar|mudiraju|mundala|murao|mushahar|muskhan|muthuraja|n[āa]dar|nagarchi|naik|nair|nayadi|nay[au]du|nuniya|obc|odd[ēe]|olgana|ontari|osh?wal|p[āa][ṭt][īi]d[āa]r|padhar|pagadai|paidi|pambanda|pamidi|panan|panchama|pannadi|paraiyan?|parathar|parava[nr]|parayan|parayi|path?aria?|pathiyan|patial|patikar|perumann[āa]n|p[īi][ñn]j[āa]r[āa]|poundra|prabhu|pulaya[nr]?|qassab|r[āa]jp([ūu]|oo)t|r[āa]ju|rab[āa]r[īi]|raja[\s\-]*mal|raidasi|raigar|rajaka|rajbanshi|ra(jkarni|njit)kar|rajwar|ramdasi|ramnami|ramoshi|rangreta|rangrez|raut|rawal|reddiar|reddy|rehar|rehgarh?|rohit|rukhi|s[āa]mbava[nr]?|s[āa]raswat|sabdakar|sadar|sadh|saharya|sakh[āa]yat|samag[āa]ra|samasi|samgar|sanhai|sansia?|sapela|sarera|sargara|sath?[vw]ara|satnami|sattada|saurashtr(a|ian)|sawasi|sebc|sedma|semm[āa]n|sengunthar|senva|settibalija|shemalia|shenva|shilpkar|shimpi|shrestha|sidhana|sinduria|siyal|sodha|solanki|sōnī|s(oo|[ūu])mr[ao]h?|s[ou]nkar|sudugadu[\s\-]*siddha|su[gk][āa]li|sunar|sundhi|sut[āa]r|suthar|sutradhar|swarankar|swarnkar|tai[\s\-]*ahom|tamadia|tamrakar|telaga|thak(ali|or|uri)|thandan|thathera|thathiar|thevar|thigala|thogata|thummagunta|t[īi]yar|tirgar|tuladhar|turha|tyagi|udayar|ujia|untwal|uppara|v[āa]ghar[īi]|v[āa][ñn]jh[āa]|v[āa]r(iy?|y)ar|vadama|vaddar|valaiyan|valhar|valluvan|valmiki|vanjari|vankar|vannan|vannar|vanniyan|v[ēe]ttuvan|velama|vēla[nr]|vell?[āa]la[nr]?|vettiy[āa]n|vishnoi|vishwakarma|[vw]?[ao]kkaliga[rn]?|vysya|w[aā]rriar|waddar|wanjara|wanjari|watal|woddar|yadav|yaithibi|yelama)(\]*s\b|[\s\-\[\]()]*(\|[^\[\]|]*\]\])?(\s*\([^().]+\))?(,\s*an?)?[\s\-\[\]]+(buddhist|cast|christian|clan|class|(communit|famil|categor)(y|ies)|descent|gotra|group|hindu|individual|j[aā]ti|jaa?tt?s?\b|khap|muslim|origin|people|person|population|quota|rajput|reservation|sikh|tribe|varna|voters|(wo)?m[ae]n))|\bchetties\b- This matches
[[reddy |something]] (something), a community, for example.
- This matches
- terms that need context words (and pluralization isn't enough to disambiguate):
\b(achari|agri(?!-)|b[ae][dr]ia|bag?di|balai|bania|bazigar|behna|beldar|bhavsar|bhulia|billava|boria|boya|bunt|bunts|burud|chakyar|chasa|chauhan|chetty|chhipa|chimpa|daivadnya|damai|darai|deha|desai|devadiga|devanga|devendrakulam|dheevara|domar|domba|gada|garha|golla|gopa|goud|gsb|hall?ikk?ar|heri|holar|jogi|kakan|kalladi|kami|kapu|karana|kharwa|koch|koda?ma|k[ōo]ta|kori|kurmi|labba|lubana|lubbe|madari|mal[ao]|mang|mannan|matang|meena|mirgan|mistri|mistry|mochi|moger|nador|naidu|nat|nayak|nayaka|niari|padanna|padmasali|palikkar|pallan|pallar|pambala|pani?kh?a|paraya|pasi|patel|patni|patua|patwa|perna|puran|rabidas?|rajak|reddi|regar|relli|ror|ruida|saini?|saliya|salvi|sanei|sansoi|santia|sapari|sapera|sarak|sarare|sardar|sarde|sarki|satani|sidh|s[ōo]n[īi]|sunga|sunri|tanla|tant(ee|i|ubai|y)|taral|tatin?|teli|thakarda|thori|thoti|toori|t[ūu]r[īi]|vargi|v[eē]la[nr]|vetan|w[aā]rrier|zampa?da)([\s\-\[\]()]*(\|[^\[\]|]*\]\])?(\s*\([^().]+\))?(,\s*an?)?[\s\-\[\]]+(buddhist|cast|christian|clan|class|(communit|famil|categor)(y|ies)|descent|gotra|group|hindu|individual|j[aā]ti|jaa?tt?s?\b|khap|muslim|origin|people|person|population|quota|rajput|reservation|sikh|tribe|varna|voters|(wo)?m[ae]n))
- wikilink targets:
\[\[(aheria|balai|behna|beldar|bhavsar|bhulia|billava|bunt \(community\)|chakyar|charodi \(community\)|chettiar|chhipi|daivadnya|damai|deha|devadiga|devanga|devendrakulam|dhobi|garha|goud|gu[rj]jar|iyer|jaa?t|kayastha?|koer[iy]|kurmi|kushwaha|labana|labbay|madari|mahishya|meena|mochi \((hindu|muslim|sikh)\)|niari|pallar|panika|paraiya|pat[uw]a|rajput|reddy|regar|relis|ror|saini|saliya|sapera \((hindu|muslim)\)|sidh \(community\)|tanti|teli|thori \(tribe\)|vaalmiki|variar)(\]\]|\|)
- misc:
backwards?[\s\-]*class?|(educationally|socially|other)[\s\-]*backward|untouchability|\b(?-i:SCs?([\s\-\/\&,]|and|or)+(ST|OBC)s?|STs?([\s\-\/\&,]|and|or)+(SC|OBC)s?|OBCs?([\s\-\/\&,]|and|or)+(SC|ST)s?)
And here's an expanded list of diffs: [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198].
Cadddr (talk) 01:17, 1 July 2026 (UTC)
- Pinging Daniel Quinlan and PharyngealImplosive7 in case you didn't see this. Cadddr (talk) 13:45, 4 July 2026 (UTC)
- @Cadddr:
Filter created and
Monitoring... – PharyngealImplosive7 (talk) 04:58, 5 July 2026 (UTC)
- Great, thanks! Since it's common for people to make caste-related edits to lines that already include caste-related terms, I'm wondering if maybe we could change these kinds of expressions:
( added_lines irlike pattern & !(removed_lines irlike pattern) )
- to:
( added_lines irlike pattern & count(pattern, added_lines) != count(pattern, removed_lines) )
- That would also require putting
(?i)at the beginning of each of the patterns. Cadddr (talk) 00:42, 6 July 2026 (UTC) - I'm wondering if it might also be a good idea to check the edit summary against the regex as well (e.g.,
summary irlike pattern). Some of the example diffs I listed above don't have enough context for the regex to be sure that they're talking about the caste, but the edit summaries give it away. Cadddr (talk) 15:46, 6 July 2026 (UTC)- Are you suggesting that we add a check that at least one term is in both the edit summary and edit or could be in either or both (and vs or check) – PharyngealImplosive7 (talk) 16:58, 6 July 2026 (UTC)
- I'm suggesting an "or" check. The filter would still trigger in all of the cases when it currently does, but it would also trigger if the regex matches the edit summary. (The rest of my comment was just an explanation of why I think this would be useful.) Cadddr (talk) 18:20, 6 July 2026 (UTC)
( added_lines irlike pattern & count(pattern, added_lines) != count(pattern, removed_lines) ) | summary irlike pattern |
- I'm suggesting an "or" check. The filter would still trigger in all of the cases when it currently does, but it would also trigger if the regex matches the edit summary. (The rest of my comment was just an explanation of why I think this would be useful.)
- Are you suggesting that we add a check that at least one term is in both the edit summary and edit or could be in either or both (and vs or check) – PharyngealImplosive7 (talk) 16:58, 6 July 2026 (UTC)
- @Cadddr:
Warn on adding redlinked categories
[edit]Per WP:REDNOT. I think this can be done using the new_html variable though it would require some testing. * Pppery * in solidarity 15:07, 26 June 2026 (UTC)
Log AI placeholder comments in infoboxes
[edit]- Task: Log and/or tag edits that contain phrases virtually only used in AI-generated infobox placeholder comments, currently "file name here"
- Reason: One of the characteristics of AI-generated articles are infoboxes with placeholder comments telling the user to add a picture, logo, etc. if one is available. It's so prevalent that it's most likely the reason why the token "Add" (capitalized) is close to the top of my list of tokens more common in AI-generated vs. human-written Wikipedia text (unless you exclude articles with infoboxes). It's tricky, though, since over the years all kinds of weird stuff and comments have shown up in infoboxes, and some of it resembles this indicator. (For instance, the comment "Add spouse if reliably sourced" is common pre-AI boilerplate).
- But after going through the search results in articlespace and draftspace, I am confident in saying that the phrase "file name here" A) basically never appeared prior to 2024, B) correlates with articles having other AI indicators, and C) is still appearing as of last week. I think "filename here" might be similar, but I haven't gone through it yet. I haven't found anything else without a lot of false positives yet.
- Diffs: Pick basically anything from the above searches
Gnomingstuff (talk) 22:47, 3 July 2026 (UTC)
- update: "filename here" is not as clear-cut, I think a lot of the stuff there is probably AI (and it is, I believe, all post-2023) but I don't have much solid to go on. "file name here" much more consistent
- there are other possible phrases I'm testing, but filtering out false positives is tricky with this stuff Gnomingstuff (talk) 18:12, 5 July 2026 (UTC)
- running list: insource:"if verified" Gnomingstuff (talk) 23:06, 5 July 2026 (UTC)
- any update on this? Gnomingstuff (talk) 07:59, 15 July 2026 (UTC)
- Once again, does anyone want to do this or respond to this at all Gnomingstuff (talk) 01:23, 20 July 2026 (UTC)
- Task: Disallow making many reports on EFFPR from one account in a short time span.
- Reason: Recent disruption from a well-known LTA.
- Diffs: special:diff/1362835643, special:diff/1362835583, special:diff/1362835556, special:diff/1362834592, special:diff/1362834567, special:diff/1362834531. Many more on special:history/Wikipedia:Edit_filter/False_positives/Reports. sapphaline (talk) 14:24, 6 July 2026 (UTC)
- I'm not sure this is worthwhile, but I'll take a look at this. Daniel Quinlan (talk) 16:53, 6 July 2026 (UTC)
Date changing filter
[edit]- Task: Catch changes of a year (a four-digit number, usually in the range 1500-2030) to another four-digit number, or a month/date to another month/date, without a source added. The filter should only apply to editors with under 100 edits, and edits with a summary of "revert"/"Rv"/"vandalism"/etc. should be exempted. Of course, since reverting a date change also produces a date change, this should be a log-only or log/Tag filter.
- Reason: This is a common form of drive-by vandalism, and is often hard to detect due to the low ORES scores and diff numbers. I have requested this filter before, but the discussion was archived before the EFMs got to it.
- Diffs: [199], plus at least twenty others I have recently reverted.
Somepinkdude (talk | contribs), in solidarity 20:37, 7 July 2026 (UTC)
Changes to filters 31 and 135 and 1233
[edit]Make EFFPR exempt from these filters as for some reason I see them triggering in the abuse log of WP:EFFPR. The only disallowing filter on EFFPR should be filter 707 as this is the page to report false positives. --Prothe1st (leave me a message)-- 14:06, 9 July 2026 (UTC)
- These filters would still prevent gibberish vandalism and the like, such as Special:AbuseLog/44248386 and Special:AbuseLog/44143039. Codename Noreste (talk • contribs) 03:18, 10 July 2026 (UTC)
- @Codename Noreste: Filters like those can still have false positives, such as Special:AbuseLog/44618506. For a page that's for reporting false positives, it's much better to be on the safe side. That's why pretty much every disallowing filter is exempt from WP:EFFPR. In cases where they make gibberish vandalism, we can just simply revert it instead. There's about 3000 people watching that page and a bunch of people who patrol recent changes. --Prothe1st (leave me a message)-- 10:17, 10 July 2026 (UTC)
- That false positive was because, after debugging, of a possible zero-width space between the sentence and the signature. Codename Noreste (talk • contribs) 17:41, 10 July 2026 (UTC)
- @Codename Noreste: Filters like those can still have false positives, such as Special:AbuseLog/44618506. For a page that's for reporting false positives, it's much better to be on the safe side. That's why pretty much every disallowing filter is exempt from WP:EFFPR. In cases where they make gibberish vandalism, we can just simply revert it instead. There's about 3000 people watching that page and a bunch of people who patrol recent changes. --Prothe1st (leave me a message)-- 10:17, 10 July 2026 (UTC)
Discussion at Template talk:Editnotices/Page/Wikipedia:Teahouse § Edit request on 11 July 2026 (add a notice regarding users to not post personal information)
[edit]
You are invited to join the discussion at Template talk:Editnotices/Page/Wikipedia:Teahouse § Edit request on 11 July 2026 (add a notice regarding users to not post personal information). There is a request to create a filter to warn people disclosing personal information like a phone number/email address at the Teahouse. Sdkb talk 14:10, 14 July 2026 (UTC)
- Before we add a separate filter that would itself increase oversight workload, could we see how often this happens with the new edit notice in place? We can wait, say 10 days, and evaluate the frequency of email addresses and phone numbers being posted before vs. after. This will need to be done by an oversighter due to suppression, of course, but I'm willing to do this. Daniel Quinlan (talk) 19:34, 14 July 2026 (UTC)
- Mr. @Daniel Quinlan, I have also just wondered for the test; can suppression be reversed? Just want to know because revision deletion can be reversed. SimpleObjects-9ei 🏖️/☀️/🥵 22:53, 14 July 2026 (UTC)
- Oversighters are able to access suppressed edits and logs, but suppression is generally not undone unless there's a mistake or something like that. Daniel Quinlan (talk) 03:33, 15 July 2026 (UTC)
- Commenting here as well: this is a large problem on talk pages as well due to people confusing the talk page with a line of communication with the article subject. I don't have statistics on this but I have personally removed hundreds if not thousands of them (many of which are too old to be reasonably oversighted, though). Gnomingstuff (talk) 21:09, 17 July 2026 (UTC)
- Oversighters are able to access suppressed edits and logs, but suppression is generally not undone unless there's a mistake or something like that. Daniel Quinlan (talk) 03:33, 15 July 2026 (UTC)
- Mr. @Daniel Quinlan, I have also just wondered for the test; can suppression be reversed? Just want to know because revision deletion can be reversed. SimpleObjects-9ei 🏖️/☀️/🥵 22:53, 14 July 2026 (UTC)