🇮🇷 Iran Proxy | https://www.wikipedia.org/wiki/Wikipedia:NEWCSD
Jump to content

Wikipedia talk:Speedy deletion

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
(Redirected from Wikipedia:NEWCSD)

See also Wikipedia talk:Speedy deletions for discussions which took place in 2006-07 before it was redirected here.

RFC: Updating T5 to account for parent templates that have been merged at TFD

[edit]
There is no consensus for this proposal. Extraordinary Writ (talk) 01:07, 22 November 2025 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.


Should T5 be updated to include the following language:

  • This applies to any and all unused subtemplates of a template that has been merged as a result of WP:TFD

Zackmann (Talk to me/What I been doing) 00:26, 23 October 2025 (UTC)[reply]

Background

Recently there have been a large number of merges via WP:TFD that have resulted in unused subpages. One such example that comes to mind was {{Infobox Twitch streamer}}. Once that merge was completed, the main template was redirected, but that left the /sanbox, /testcases and /styles.css pages. A user attempted to speedy delete them, but that was rejected by an admin because WP:SD does not include "subpages of merged templates". Thus for these unused templates were forced to file a full TFD.

Now I would argue there are other criteria that fit...

But given that there are a number of admins who have felt in the past that these types of pages do NOT fit T5, I think we should spell out the fact that they do to make it clear. I have not heard ANYONE say that these types of pages should not be deleted. It has been a purely an objection on a technicality. Zackmann (Talk to me/What I been doing) 00:26, 23 October 2025 (UTC)[reply]

Discussion on updating T5

[edit]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

RFC: Time to promote or repeal WP:X3?

[edit]
The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.
The result of the discussion is that there is a general consensus to mark X3 as obsolete. Among those with this preference there is a general consensus that the language of R3 will cover new pages being created with this issue, and any stragglers can be dealt with in the usual ways. Primefac (talk) 10:50, 29 November 2025 (UTC)[reply]

It has been over a year since the temporary criterion WP:X3 was enacted. At present, it looks as though the backlog of titles which this criterion applies have now been deleted. (Further details in the following comment.) At this point, should we make this criterion "Obsolete", promote this criterion to a permanent criterion (would be "R5"), do nothing to the criterion at the present time, or take some other action? Steel1943 (talk) 19:05, 29 October 2025 (UTC)[reply]

Additional information: At this point, most likely 99% of the titles on Wikipedia that do not have a space before a "(" do not contain a disambiguator. (Examples: Chemical names, purposeful parenthesis in the title, etc.) After personally resolving issues with titles found by a intitle:/\(India\)\(/" search (locating titles with "(India)(" in them) during the previous few months, as well WP:X3 deletions in response to titles tagged by Helpful Raccoon this month, in addition to all other editors who have tagged and/or deleted applicable redirects during the past year, it could be true that any titles that currently exist without a space before a "(" could be considered false positives (ineligible) for WP:X3 speedy deletion. Steel1943 (talk) 19:07, 29 October 2025 (UTC)[reply]
  • Promote to permanent criterion R5. Given the extensiveness of the amount of discussions in the past, I believe it could be helpful to retain this criterion on a permanent basis. Steel1943 (talk) 19:07, 29 October 2025 (UTC)[reply]
    • I'm confused. You state that almost all titles that could be deleted by this title have been deleted, which is an argument for marking it obsolete (c.f. WP:NEWCSD point 3) yet you want to make it permanent. Why? Thryduulf (talk) 20:18, 29 October 2025 (UTC)[reply]
      It's the whole "if it could happen once, it could happen again" problem. Speedy deletions for this criterion will become reactive instead of proactive and will happen very seldom (basically, only when new titles are created), such as what eventually occurred shortly after WP:R4 was enacted. That, and comparing the other temporary criteria, new titles will not be created for WP:X1 anymore since the respective editor is indefinitely blocked, and WP:X2 eventually resulted in a solution that made the criterion essentially useless, including the consideration that the criterion only applied to titles that were created during a specific timeframe. The difference here is that titles eligible for WP:X3 could still be created in the future (unless somehow we manage to create a title creation blacklist entry properly, but I do not see that happening due to several possible false positives.) Steel1943 (talk) 20:23, 29 October 2025 (UTC)[reply]
    For what it's worth, the concern that I have with merging this criterion into R3 is that R3 contains the requirement of the redirect being "recently created". There should be no "recently created" restriction for redirects that are currently eligible for X3. Steel1943 (talk) 21:44, 29 October 2025 (UTC)[reply]
    That is indeed a good point, and in my opinion, the most important question in this discussion. However, if we don't have large-scale cleanup to do, X3/R5 might fail criterion 3 ("frequent") as most cases would have likely been caught by R3. Beyond that, I am not familiar enough with the history of R3 to state if the reasons for including a "recently created" provision would also apply to X3 or not. Chaotic Enby (talk · contribs) 21:51, 29 October 2025 (UTC)[reply]
    R3's time limit limits the potential damage from misinterpration, as actually happened at e.g. Central Reorganization Committee, CPI(ML)Central Reorganization Committee, CPI(ML) (created in 2004!), Albatros B.I(Ph)Albatros B.I(Ph), and Public Offices (Candidacy and Taking Up Offices)(Miscellaneous Amendments) Bill 2021Public Offices (Candidacy and Taking Up Offices)(Miscellaneous Amendments) Bill 2021. —Cryptic 22:10, 29 October 2025 (UTC)[reply]
    Yeah, that's the oddness about all of this. I mean, I can provide an argument that basically states that "frequent" may not be a valid criteria to establish a speedy deletion criterion anymore. Best example I can probably provide is WP:R4: It was previously bundled in with WP:G6, but was separated after determining that giving it its own criterion since it specifically targets redirects was the best option. Quite frankly ... I'm almost certain that there has not been a page tagged for WP:R4 for quite a while ... I would be shocked if there was even a single page tag during the previous year. Steel1943 (talk) 00:00, 30 October 2025 (UTC)[reply]
@BodhiHarp, Cryptic, and Chaotic Enby: Pinging participants of #Renaming X3 after realizing the potential redundancy of this discusion. (That, and I noticed that Chaotic Enby presented yet another option: Essentially merging the criterion with WP:R3.) Steel1943 (talk) 20:47, 29 October 2025 (UTC)[reply]
  • New ones are redundant to R3, and the (lack of) wording in X3 makes it error-prone. Recent relevant discussion on my talk page. The merge to R3 isn't a new option; it was discussed (at some length, IIRC) during X3's original enaction. I don't think R3 needs any change to incorporate this, either, though if that's a sticking point we can add a couple words to the effect of "including malformed disambiguators" to it. I haven't investigated whether the assertion that all the old ones have been dealt with is correct, though have no reason to doubt it. Do not promote, in any event. —Cryptic 21:07, 29 October 2025 (UTC)[reply]
    I feel it would help to mention it. AFAIK the consensus around R3 is that redirects normally only qualify as implausible where there are multiple typos in the title. Malformed disambiguators typically have only one typo. (Though, like many speedy criteria, the multiple-typos idea is hardly how it gets applied in practice.) J947 ‡ edits 23:31, 29 October 2025 (UTC)[reply]
  • Merge (i.e. mark as obsolete and clarify R3) per my argument in the previous discussion. New cases already fall under R3, but the wording could make it more explicit. No need to make it a permanent criterion as it would essentially be redundant. I like Cryptic's suggestion for added wording. Chaotic Enby (talk · contribs) 21:12, 29 October 2025 (UTC)[reply]
  • Merge to R3 (when clarified). The issue is more then likely to return but can be handled on a case by case basis by R3. The Banner talk 21:32, 29 October 2025 (UTC)[reply]
  • Repeal. As someone who opposed this criterion's creation, I maintain that this was an unnecessary project because the redirects in question were harmless and should've been left alone rather than deleted. -- Tavix (talk) 03:02, 30 October 2025 (UTC)[reply]
  • Merge to G6 ({{db-error}}), or failing that, R3 with clarified wording (per The Banner). A missing space before parenthesis still a plausible mistake to make – so IMO it's debatable whether this would clash with the current language of R3 – but it is very much an error for which speedy deletion has been justified, as per numerous RfDs and the enactment of X3. Complex/Rational 16:44, 30 October 2025 (UTC)[reply]
  • When I first proposed creating an X-series of speedy deletion criteria, they were meant to be solutions to specific problems. X-criteria need to be constrained to an enumerable set of articles so that when that set has been worked through, the criterion can be retired. Therefore X3 can't continue in its current form and should be repealed.—S Marshall T/C 22:57, 4 November 2025 (UTC)[reply]
  • Retire. Any new instances that are redundant to R3 can be deleted under that criterion, any that aren't shouldn't be speedily deleted. There is no need to make any changes to R3. Thryduulf (talk) 11:50, 5 November 2025 (UTC)[reply]
  • Retire per Thryduulf. Very open to a revisit if you can show pages slipping through the cracks and gumming up RfD, but let's try the planned process first Tazerdadog (talk) 21:49, 5 November 2025 (UTC)[reply]
  • Comment: I just deleted a redirect without a space, which had been tagged as R3. Given this, it's very possible the statistics on X3 use are inaccurate. Because I had seen this discussion earlier today, I deleted as X3 instead. This may be more common than realized. Significa liberdade (she/her) (talk) 01:00, 7 November 2025 (UTC)[reply]
    I made Quarry 98877 to look for instances of this, and (discarding seven article deletions) got 42 deletions, of which 26 were R3 and 16 were G6. I excluded many things from this query, like R2, {{db-move}}, G7, G8, and the A-series criteria, since I think X3 would be inapplicable, unreasonable, or moot in those cases. I also made a separate query for the RfDs (Quarry 98880) and got 50; since RfD deletion summaries don't usually include the rationale, a full analysis is beyond the scope of this comment. jlwoodwa (talk) 02:22, 7 November 2025 (UTC)[reply]
    I tagged that redirect, Unionville Meadows(Public School), as R3 because of this discussion. As far as I can tell, that was the only X3-eligible page created within the past 4 weeks. Helpful Raccoon (talk) 03:23, 7 November 2025 (UTC)[reply]
    It may not be that my estimation was inaccurate, given I was using a list of titles generated in the early-mid 2024, but rather that these titles are still being created. For example, Unionville Meadows(Public School), the redirect mentioned by Helpful Raccoon above: It was created about a week or two ago. Steel1943 (talk) 15:34, 8 November 2025 (UTC)[reply]
  • Merge to R3 It seems like the best way to enact this permanently, which it should be. ᴢxᴄᴠʙɴᴍ (ᴛ) 22:27, 7 November 2025 (UTC)[reply]
  • Repeal X3 per Thryduulf and to a lesser extent Tavix. If these redirects are recent enough, then they qualify for WP:R3 (which I imagine will be the majority of cases as they are being closely tracked by a small number of attentive editors). Otherwise, they should be treated the same as most other single-typo redirects (e.g. those with double opening parentheses in their qualifiers) and either individually brought to RfD or even ignored altogether. Redireditor (talk)aka Dsuke1998AEOS 01:26, 9 November 2025 (UTC)[reply]
  • Merge to R3 - I'll miss the X category, but this really needs to be enacted permanently. Z E T A3 02:16, 9 November 2025 (UTC)[reply]
    My understanding is that the X category would remain after X3's repeal albeit with no current criteria. This allows us to spin up an X4 criterion if a need arises. Tazerdadog (talk) 21:21, 12 November 2025 (UTC)[reply]
Repeal. No longer worthy of an extraordinary criterion. Also, X3 is somewhat redundant to R3.
Whether R3 needs clarifying (and whether it should be clarified include or exclude these redirects) could be discussed afterwards. ~2025-33069-57 (talk) 14:32, 12 November 2025 (UTC)[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

U5 discontinuation

[edit]

Recently heard about the discontinuation of U5. I'd like to ask: what will happen to user pages previously deleted under that criterion? I found nothing about that at Wikipedia:Replacement of CSD U5 FAQ. Thanks, 1isall (he/him) (talk | contribs) 18:04, 4 November 2025 (UTC)[reply]

Nothing; they stay deleted. CSD repeals are not retroactive.
We could probably modify WP:REFUND to allow requests for U5-deleted stuff to be undeleted, though. Not sure if that's worth the effort, as I doubt there'd be many requests. Elli (talk | contribs) 18:11, 4 November 2025 (UTC)[reply]
As an undeleter, I would expect a good reason for the requestor to have a hope of returning that sort of page. Normally inappropriate pages would not be restored (eg someone's homework, or creative writing) but some undeleters will email the page to the requestor. Graeme Bartlett (talk) 20:10, 18 November 2025 (UTC)[reply]

Edit request 11 November 2025

[edit]

Please add the following to WP:CSD R2 and WP:XNR:

  • <onlyinclude>A rare expample is {{no redirect|Hatnote|this}}, which redirects to [[Wikipedia:Hatnote]]. Redirects to major internal self-reference pages are appropriate and useful for readers and new editors not familiar with Wikipedia's namespaces.</onlyinclude>
  • Per [[WP:CSD R2]], ''with the exception'' of the '''Category:''', '''Template:''', '''Wikipedia:''', '''Help:''', and '''Portal:''' namespaces, redirects from the main (article) namespace to any other namespace are [[WP:SPEEDY|speedily deleted]]. {{#section:Wikipedia:Speedy deletion|R2. Cross-namespace redirects}}

--~2025-32430-55 (talk) 19:57, 11 November 2025 (UTC)[reply]

 Not done for now: please establish a consensus for this alteration before using the {{Edit extended-protected}} template. NotJamestack (talk) 20:48, 11 November 2025 (UTC)[reply]
[edit]

WP:G15 says, in part:

  • Implausible non-existent references: This may include external links that are dead on arrival, ISBNs with invalid checksums, and unresolvable DOIs. Since humans can make typos and links may suffer from link rot, a single example should not be considered definitive. Editors should use additional methods to verify whether a reference truly does not exist.
  • Nonsensical citations: This may include citations of incorrect temporality (e.g., a source from 2010 being cited for a 2022 event), DOIs that resolve to completely unrelated content (e.g., a paper on a beetle species being cited for a computer science article), and citations that attribute the wrong author or publication.

Does "citations that attribute the wrong author or publication" belong with the previous category, as |author=Alice|title=Bob's Book is an implausible non-existent source?

Or is the first one supposed to be all about the URLs and id numbers? WhatamIdoing (talk) 00:51, 15 November 2025 (UTC)[reply]

I think it's fine how it is. -- LWG talk 01:52, 15 November 2025 (UTC)[reply]
My reading of the distinction is that the first of these bullet points is for citations which don't go to a document at all; while the second is for citations that do, but not one which matches what the citation says. So
  1. ^ Alison, Alice. Bob's Book. Generic Press, 2025.
is an example of an "implausible non-existant reference", while
  1. ^ Alison, Alice. Bob's Book. Generic Press, 2025.
is an example of a "nonsensical citation". Something without a link like
  1. ^ Alison, Alice. Bob's Book. Generic Press, 2025.
might be either, depending on whether Generic Press ever published a Bob's Book by Alice and, if so, whether it has anything whatsoever to do with what's being cited to it. —Cryptic 02:17, 15 November 2025 (UTC)[reply]
This might be a stupid & pedantic question, but after reading this explanation, I put citations w/ example.com (or similar) URLs in the "implausible" category then, right? I've been double checking that the article/book doesn't appear to exist just in case somebody leaves multiple fictitious URLs in by mistake, but I wasn't entirely sure which selection to pick in the drop down and have instead been selecting one arbitrarily and explaining my thought process/confirmation steps at length in the explanatory notes box. (So sorry admins who process my G15s!) GreenLipstickLesbian💌🦋 02:49, 15 November 2025 (UTC)[reply]
(edit conflict) For clarity (as they appear identical when rendered) the link in @Cryptic's first example leads to https://en.wikipedia.org/w/broken.php (i.e. a broken link) while the second links to https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Speedy_deletion
The first source does not exist, and it is implausible for it to exist - for example if Generic Press published a book by Alice called Bob's Brook it is plausible that the article writer typoed and this limb of the criterion is not met (the source should be evaluated against the second limb, and the typo fixed if it turns out to be plausibly relevant to the article content).
The second source exists, but it is both irrelevant to the subject of the article and implausible that it could ever be relevant. For example an academic paper about a beetle species is almost certainly implausibly relevant to an article about a computer science topic (although what claim it is being used to support might make a difference), but it is plausibly relevant to an article about a different beetle species even if that species isn't (currently) mentioned in the article. Thryduulf (talk) 02:49, 15 November 2025 (UTC)[reply]

Template:Db-notwebhost-deleted

[edit]

Since WP:U5 was removed, Template:Db-notwebhost-deleted should either be updated or sent to TfD. Gonnym (talk) 17:33, 15 November 2025 (UTC)[reply]

WP:FAKEARTICLE should also be updated as it links to the deleted U5 template. Gonnym (talk) 17:35, 15 November 2025 (UTC)[reply]
{{Db-u7-deleted}} has nearly-identical wording to that template, with the exception of the 'not a free web host' clause, so {{Db-notwebhost-deleted}} could be merged into it. Xeroctic (talk) 12:32, 17 November 2025 (UTC)[reply]
It should not be deleted, as that would disrupt all the historical uses. But it could be deprecated. Graeme Bartlett (talk) 20:06, 18 November 2025 (UTC)[reply]
It's a subst-only template, so changing or deleting it won't affect existing uses. jlwoodwa (talk) 20:07, 20 November 2025 (UTC)[reply]

Issue regarding the A-series criteria

[edit]

The A-series criteria such as A7 has been misapplied to drafts. Should we fix this or should we do nothing? - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 18:47, 18 November 2025 (UTC)[reply]

@BodhiHarp: Can you provide some examples? Children Will Listen (🐄 talk, 🫘 contribs) 18:49, 18 November 2025 (UTC)[reply]
I think I saw it here, but it may have been suppressed. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 18:51, 18 November 2025 (UTC)[reply]
It was indeed suppressed, so this has nothing to do with speedy deletion. Children Will Listen (🐄 talk, 🫘 contribs) 18:55, 18 November 2025 (UTC)[reply]
I thought it was tagged as A7 when it existed. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 18:56, 18 November 2025 (UTC)[reply]
It was tagged for both A1 and A7 by user:Aesurias. Thryduulf (talk) 21:46, 18 November 2025 (UTC)[reply]
I thought I added the A1. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 21:46, 18 November 2025 (UTC)[reply]
Please don't tag oversightable material for speedy deletion, even if you think it's "only borderline". That makes it much more visible. (Facially-nonapplicable criteria like A* on non-articles even more so.) If you can't actually delete it yourself, mail oversight directly and let them deal with it. —Cryptic 21:51, 18 November 2025 (UTC)[reply]
What the admin probably did was delete the page under a CSD to attract less attention than say, "IAR", or even worse, "has private info." Children Will Listen (🐄 talk, 🫘 contribs) 03:45, 19 November 2025 (UTC)[reply]
An admin deleting it, especially with a non-attention-attracting log entry - and an A7 or A1 on a draft does stick out - is all well and good. Same situation as Wikipedia:Revision deletion#Hiding oversightable material prior to Oversight. Tagging the page puts it into the very-visible Category:Candidates for speedy deletion, sometimes for a long time, where it will be seen not just by admins (who may or may not act on it) but by, for example nonadmins looking for pages to untag, and projects that preferentially mirror pages likely to be imminently deleted. —Cryptic 04:00, 19 November 2025 (UTC)[reply]
Yes, of course. I assumed an admin deleted it before sending it to oversight, not an editor tagging it for speedy deletion. Children Will Listen (🐄 talk, 🫘 contribs) 04:10, 19 November 2025 (UTC)[reply]
Thryduulf said just above that it had been tagged, and BodhiHarp said he had tried to tag it too. —Cryptic 04:43, 19 November 2025 (UTC)[reply]
It was tagged for A1 at 05:21, for A7 at 05:24 and detagged at 05:59. The email to oversight was received at 06:02 and it was oversighted at 10:17 (which is quite a long gap). Thryduulf (talk) 10:24, 19 November 2025 (UTC)[reply]
I think A10 has the most support for applying to drafts. There is no value in having a draft on an article which already exists, and these seem to be common and easily recognized. I recently encountered Draft:भास्कराचार्य, which is in Hindi and seems to be about a person who already has an article, Bhāskara II. –LaundryPizza03 (dc̄) 05:01, 1 December 2025 (UTC)[reply]
In many cases though the draft title should be redirected to the relevant mainspace article (after merging anything useful that isn't already there) rather than deleted, to discourage the creation of future redundant drafts. Thryduulf (talk) 05:07, 1 December 2025 (UTC)[reply]
I think we should redirect, not expand or change A10 to cover drafts. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions, deleted contributions) 05:20, 1 December 2025 (UTC)[reply]
Not that "easily recognized", it seems - that draft was declined, rather amusingly, on notability grounds. GreenLipstickLesbian💌🧸 05:10, 1 December 2025 (UTC)[reply]

Footnote for WP:G5

[edit]

The page currently says "Note that accepting a draft at WP:Articles for creation is not considered a substantial edit" and cites a deletion review. Have there been any other deletion reviews or related debates, or an RfC to substantiate this? Ritchie333 (talk) (cont) 11:55, 21 November 2025 (UTC)[reply]

Broaden G15 to align with new guideline

[edit]
ON HOLD
The clear consensus here is that people want NEWLLM to be stable before there's any considering of a change to G15. I'll close this rather than blank it so it can be referenced in a later RFCBEFORE for a G15 change once NEWLLM is settled. Thanks everybody for the feedback. Athanelar (talk) 21:51, 24 November 2025 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Given that WP:NEWLLM is now accepted as guideline, should CSD G15 be expanded to apply to all articles that consist primarily of AI-generated content, whether or not that content has been subject to human review? Athanelar (talk) 15:43, 24 November 2025 (UTC)[reply]

  • No / oppose. From the close, the current wording (which changed during the debate) does not enjoy a particularly strong consensus and requires further development. In particular we need community consensus on (a) How to identify LLM-generated writing and (b) How to deal with it when it does occur. Given that there is neither consensus on how to identify LLM-generated writing not consensus on how to deal with it, we should not expand the G15 criteria. Cheers, SunloungerFrog (talk) 15:53, 24 November 2025 (UTC)[reply]
    Well, what I'm essentially doing here is trying to get community consensus on (b); i.e., my answer to (b) would be that if consensus determines an article to be primarily AI-generated, then G15 should apply.
    Also, it just seems rather contradictory to on the one hand have a policy guideline say that AI shouldn't be used to generate new articles wholesale, and then on the other hand have a CSD which implies that AI-generated articles are fine to some extent as long as they've been reviewed by a human for factual accuracy etc. Athanelar (talk) 15:56, 24 November 2025 (UTC)[reply]
Seems a bit premature to consider an expansion at the moment. If there is a perception G15 isn't working well, would be better to discuss the specific deficits are, this proposal is not very specific as to what changes are to be made to the criteria, which CSD need to be. Also suggest this be converted into a regular (non-RFC) and ongoing discussion unless and until such specific recommendations have been developed. Alpha3031 (t • c) 15:56, 24 November 2025 (UTC)[reply]
@Athanelar: I think this is a discussion worth having, but I'd encourage you to not put it to an RfC just yet, and instead workshop wording either here or at WP:VPIL. The main question would be how we reliably determine whether an article is AI-generated. The "I know it when I see it" approach works fine for removing content from an article, blanking and redirecting it, or even draftifying it, because these are all easily reversible. With deletion, on the other hand, we usually want a more objective standard. G15 currently addresses this with its three categories of AI indicators, but all of those are related to lack of human review, so wouldn't be helpful for the purposes of this RfC. The way I see it, there are two obvious ways G15 could be expanded:
  • Add a clause for where the user acknowledges an LLM was used, even if they claim they reviewed the article. This would extend the WP:NEWLLM consensus without adding any subjectivity to the mix.
  • Add something flexible but still trying to be objective, like A7/9/11 and G11 have, something like "Articles that any reasonable person familiar with large language models would understand to have been generated from scratch using one"
But this is why I suggest workshopping this first in a pre-RfC setting. -- Tamzin[cetacean needed] (they|xe|🤷) 15:56, 24 November 2025 (UTC)[reply]
Thanks for the feedback. I've removed the RFC tag to use this thread to workshop the precise expansion to be later made into an RFC. Would it be alright for me to post a link to this thread at WP:VPIL to get eyes on it? Athanelar (talk) 16:00, 24 November 2025 (UTC)[reply]
As long as you word it carefully so that you don't accidentally encourage a load of "LLMs bad, mmmkay" type comments that will just get in the way of a productive discussion that will be fine. CSD criteria require very careful wording to make sure they meet all of the WP:NEWCSD criteria. Something as vague and vibes-based as the recently promoted guideline would be actively harmful here. Thryduulf (talk) 16:08, 24 November 2025 (UTC)[reply]
@Athanelar I have been working on a few proposals for expanding G15. Can we collaborate at WT:AIC before we take this elsewhere? NicheSports (talk) 16:56, 24 November 2025 (UTC)[reply]
I'm always up to discuss this sort of thing, ping me if you open a thread there with one of your proposals. Athanelar (talk) 17:01, 24 November 2025 (UTC)[reply]
Will do later today. I'll incorporate Tamzin's first suggestion above, I have considered that one as well. Thanks Athanelar! NicheSports (talk) 17:02, 24 November 2025 (UTC)[reply]
shouldn't have !voted as the RFC tag was closed. My apologies to Athanelar NicheSports (talk) 17:03, 24 November 2025 (UTC)[reply]
Oppose for now NEWLLM isn't set in stone, and there are three open proposals as of now. Think this should be closed and reopened after NEWLLM becomes more stable per the clear consensus. Z E T AC 21:37, 24 November 2025 (UTC)[reply]

Oppose. We are currently doing an WP:RFCBEFORE on these guidelines. Technical details like CSD come after the guidelines are settled. 〜 Festucalex • talk 17:05, 24 November 2025 (UTC)[reply]

Given that the current question about the guideline is how or if it should be expanded and not whether it should be reduced, the guideline will still forbid wholly/primarily AI-generated new articles in whatever form it ends up in, so my case for an expansion of G15 would remain the same either way.
This is in itself an RFCBEFORE of said expansion. The question is essentially 'do you think G15 will need to be changed given that we now have a guideline which forbids new AI articles regardless of review,' with the aim to later develop specific wording for that change if people agree it needs to be changed. Athanelar (talk) 17:14, 24 November 2025 (UTC)[reply]

Oppose The implications of WP:NEWLLM are still being worked out, but in particular we currently have no standard other than consensus for determining whether articles "consist primarily of AI-generated content", and a CSD bypasses that consensus process. I think Tamzin's ideas deserve further consideration, though. -- LWG talk 18:01, 24 November 2025 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Alter G15 to avoid tagging within minutes of creation?

[edit]

I have noticed that the descriptions of A1 and A3 include Do not tag under this criterion in the first few minutes after a new article is created. The reason for this is because the creator may still be working on the article to resolve the problem. I think G15 should have a similar caveat: 'Do not tag under this criterion in the first few minutes after a new page is created'. This is because I think it would be plausible to copy-paste the LLM text, click 'publish', then go back and do the human review, and the concern G15 addresses are fixable by the creator (by doing the review) and is not immediately harmful to the project. QwertyForest (talk) 19:02, 25 November 2025 (UTC)[reply]

I agree with this. While they would ideally solve any issues before publishing, Wikipedia benefits in whichever order they do it. Only if they don't fix the issues, after having an opportunity to do so, should we be speedily deleting their work. Thryduulf (talk) 19:39, 25 November 2025 (UTC)[reply]
+1 IMO the only criteria that should be applied immediately are G3, G10, and G12, following from the same rationale. And I also think it likely that comically bad AI slop won't be reviewed anyway, so there's no harm in waiting at least ~15 minutes to G15 it. Complex/Rational 20:30, 25 November 2025 (UTC)[reply]
You've made some good points there. Maybe the 'don't pounce on new pages immediately' should be applied to all criterion except G3, G10 and G12. Apart from those three, there's no harm in waiting to see if the creator fixes itand, even if they don't, waiting a short while might make tagging a bit less bite-y. QwertyForest (talk) 21:19, 25 November 2025 (UTC)[reply]
I'm not sure about this now that WP:NEWLLM has passed. If the first revision was G15-able then the next will probably fail at AfD as an obvious violation of "don't use LLMs to write an article from scratch". lp0 on fire () 13:31, 29 November 2025 (UTC)[reply]
Firstly we don't know that. Secondly, if subsequent revisions solve all the actual problems (or at least the significant majority of them) then it seems unlikely that it will be deleted at AfD. This is especially true if the problems are one of style rather than of verification. Thryduulf (talk) 11:31, 30 November 2025 (UTC)[reply]
I can't disgree with that, but I think it's be worth waiting until it's a bit clearer how WP:NEWLLM is going to be applied. lp0 on fire () 14:10, 30 November 2025 (UTC)[reply]

U6 ineligible template

[edit]

I have enough vague memories of seeing SPAs collaborate in sandboxes for editathons and classes, then one participant/account copies the material to a main space article. The need to preserve edit history, in these cases, trumps U6 IMO, so I whipped up a super basic template: {{U6 ineligible}}, somewhat analogous to {{Promising draft}}. This should hopefully mark to admins/patrollers that these pages shouldn't be procedurally deleted.

But anyways, if anybody wants to fiddle with the wording, please do, tell me anything I've missed, nominate it for TfD, please feel free. I'm useless at templates, template categories, and this sort of behind the scenes writing, so I know there's plenty of room for improvement. I did debate expanding the scope to include cases where the page creator has significant global contributions, the page is being used by other editors, or even just 'another editor thinks the speedy would be controversial so pleas take to MfD if you want this vanquished', but I wasn't entirely sure how to word that, and I didn't want to risk muddying the water too much.and I don't trust myself to make the parameters work. GreenLipstickLesbian💌🧸 09:58, 27 November 2025 (UTC)[reply]

In the first case, shouldn't such a thing be fixed as described in Wikipedia:Fixing cut-and-paste moves? In other cases, if a page is that useful then I wonder if it should be moved somewhere out of the userspace of a low-activity user. Anomie⚔ 15:02, 27 November 2025 (UTC)[reply]
Maybe? Sometimes if the pages have dual history (which is where I've seen this happen most, because you can't use the 'move' button), it's not always practical. GreenLipstickLesbian💌🧸 17:27, 27 November 2025 (UTC)[reply]

Allow creator to remove U6 tags?

[edit]

U6 is intended to be a fairly mechanical process, similar to G13 for deleting expired drafts. Since the creator is allowed to remove G13 tags, should U6 be the same? Anomie⚔ 17:55, 29 November 2025 (UTC)[reply]

I would think so; part of the definition is that it has not been edited in 6+ months, so if edits are made to the page (e.g. removing the tag) then it no longer meets U6. I will note the template does currently mention this. Primefac (talk) 17:35, 30 November 2025 (UTC)[reply]
Per lack of objection, I went ahead and added this. Anomie⚔ 16:48, 4 December 2025 (UTC)[reply]

Weird corner cases with U6

[edit]

I've been quietly following the implementation of criteria U6 and U7 and am very glad that (a) they only involve subpages and (b) old pages in the former category are being tagged slowly the way they are. The absolutely overwhelming majority of them in the U6 bot categories should be deleted but I found a few that shouldn't (see my relevant contribs). The most interesting by far were the pages now at User:Home Row Keysplurge/Old talk 0x0000 and User:Home Row Keysplurge/Old talk 0x0001. They were talk page archives from 2004 in userspace (which wasn't an unheard-of location for them back then) numbered in hexadecimal (which was clearly a personal touch). They were left behind after a username change in July 2005, because pages weren't automatically moved in this situation until December 2006. My #1 law of Wikipedia: there'll be weird corner cases everywhere. I'll be monitoring the U6 categories just in case (mostly checking for unusual page titles) and encourage others to do the same. I was reminded of the implementation of the bot by this bureaucrats' noticeboard thread; I should have followed the BRFA more closely. Graham87 (talk) 07:11, 30 November 2025 (UTC)[reply]

The bot should be checking if the subpage was actually created by the user in question; else stuff like everything under User:UBX will be deleted, for example. Something like this is already happening with ;~enwiki (Special:PrefixIndex/User:;~enwiki.) Children Will Listen (🐄 talk, 🫘 contribs) 07:18, 30 November 2025 (UTC)[reply]
UBX is on the bot's ignore list. As for the ;~enwiki cases, I noticed those too but didn't have any strong feelings on them. Graham87 (talk) 07:22, 30 November 2025 (UTC)[reply]
Yeah thinking about it, maybe they should be on the ignore list too, for similar reasons to the page I requested there earlier. Graham87 (talk) 07:33, 30 November 2025 (UTC)[reply]
@ChildrenWillListen: I've ended up removing their speedy deletion tags, because Extraordinary Writ added an entry in the ignore list for User:Beecat, a similar situation. Extraordinary Writ, could you add the ";~enwiki" one as well? Graham87 (talk) 05:59, 1 December 2025 (UTC)[reply]
Sure, done. Extraordinary Writ (talk) 06:03, 1 December 2025 (UTC)[reply]
@Anomie: Is there a reason the bot doesn't check the page creator? It doesn't sound like it's too difficult to implement. Also, what if someone abandons a page in their userspace and another, more experienced user started to work on it? Children Will Listen (🐄 talk, 🫘 contribs) 07:27, 30 November 2025 (UTC)[reply]
I was thinking that, but it feels a bit gameable to me to have a loophole like that. Or maybe pages where the creator is different from the user page title can be under a different category? Graham87 (talk) 07:33, 30 November 2025 (UTC)[reply]
@Graham87: If someone tries to game the system like that to keep their unencyclopedic sandboxes forever, they can be blocked per WP:SOCK and/or WP:NOTHERE. Children Will Listen (🐄 talk, 🫘 contribs) 07:36, 30 November 2025 (UTC)[reply]
Because no one ever suggested that U6 (or the bot-tagging for U6) should only apply to pages created by the user owning the userspace. I'll run some queries to see just how many pages in the backlog would be affected by this. I note detecting "creator" is surprisingly tricky to do accurately 100% of the time (see Wikipedia:Village pump (technical)#c-Anomie-20251130023400-Ktrimi991-20251130020100 for example), but simple methods can be "good enough" most of the time. Detecting "some more experienced user started to work on it" would be a lot harder, as having a bot differentiate "work on it" from "drive-by tagging, linter cleanup, etc" would be difficult. Ideally, if someone wants to work on such a page, they'd move it to draftspace or something instead of leaving it as a subpage of a disappeared user. Anomie⚔ 15:45, 30 November 2025 (UTC)[reply]
I don't think this would be a very good exclusion. Having made a list (1253 pages total) and spot-checking a few entries, we seem to have quite a few cases where an instructor or someone created sandboxes for Education Program students, cases where an instructor had their students create pages in the instructor's userspace, some cases of WP:REFUND that apparently never went anywhere (and were somehow done without the owning user actually posting at WP:REFUND?), cases where someone moved a draft article from a userpage to a subpage, and so on. Anomie⚔ 16:24, 30 November 2025 (UTC)[reply]
@Anomie: Thanks for checking this out. I agree there's no real reason to preserve these drafts, and there are plenty of unencyclopedic content lurking in there (example). Children Will Listen (🐄 talk, 🫘 contribs) 17:51, 30 November 2025 (UTC)[reply]
Thanks for helping monitor these! The week delay before the pages actually become eligible for deletion is specifically intended so humans have time to look for anything that should be kept, whether things like these old archives, promising drafts, or anything else. Anomie⚔ 15:45, 30 November 2025 (UTC)[reply]
No worries. I've written more ramblings at User talk:CSD U6 Bot/Ignore list#Making this page template-protected, etc.? Graham87 (talk) 07:01, 1 December 2025 (UTC)[reply]

6 months criteria

[edit]

i deleted User:Qazphynux/sandbox in error before I noticed the 6 months criterion. The content was clearly inappropriate so I blanked it. I think the 6 months is not helpful with nonsense like this. Doug Weller talk 19:10, 1 December 2025 (UTC)[reply]

Why is it harmful though? Unless you can clearly articulate some actual (not just theoretical) harm that leaving that in userspace for 6 months is causing then speedy deletion is obviously inappropriate. Thryduulf (talk) 22:02, 1 December 2025 (UTC)[reply]

Proposal for A11

[edit]
Withdrawn. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 17:54, 7 December 2025 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Shouldn't A11 be a general criterion? There is no direct speedy deletion criterion for invented stuff in the draft space. We have A11, but that's mainspace articles only, and we have G3, but that's hoaxes, not inventions that are indicated to be invented.

We may want to either:

  • Replace it with a general criterion G16
  • Or merge to G3

Also, this is similar to how G12 was formerly A8. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions, deleted contributions) 05:19, 2 December 2025 (UTC)[reply]

How often does this get nominated at MfD? Is it always deleted? Obviously we can't just extend A11 to be a G criterion without modification as there is an allowance for a limited amount of invented stuff in userspace and project space (see for example many of the Wikipedia:Department of fun subpages). Merging with G3 would be a terrible idea because that's for bad-faith creations and a significant proportion of A11 stuff is people contributing in good faith but simply not understanding what Wikipedia is for. Thryduulf (talk) 14:03, 2 December 2025 (UTC)[reply]
Please explain how the 4 criteria noted at the top of the talk page and the edit notice are met. In particular, what research into frequency have you done? -- Whpq (talk) 14:34, 2 December 2025 (UTC)[reply]
FWIW, U7c does apply WP:NOTMADEUP to userspace, but only under a limited set of circumstances. I'm not sure it would be desirable to apply it any more broadly than that in userspace. -- Tamzin[cetacean needed] (they|xe|🤷) 08:03, 7 December 2025 (UTC)[reply]
No. Inventions and discoveries can be perfectly good drafts, and the editor adding the content having a WP:COI is a reason to have it drafted in userspace or draftspace, and is not a reason to delete it from userspace or draftspace. SmokeyJoe (talk) 10:54, 7 December 2025 (UTC)[reply]
See this. - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 17:25, 7 December 2025 (UTC)[reply]
And? We have one example of something that can be deleted using an existing speedy deletion criterion, suggesting that WP:NEWCSD point 4 isn't met. Thryduulf (talk) 17:48, 7 December 2025 (UTC)[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

X3 category

[edit]

I notice we still have Category:Candidates for speedy deletion as redirects with no space before a parenthetical disambiguation. Redirect to R3? - BᴏᴅʜıHᴀᴙᴩ (talk, contributions) 21:05, 8 December 2025 (UTC)[reply]

I'd say take it to CfD. Pages should generally not be being added to speedy deletion categories directly, and anyone still nominating pages under X3 they need to stop and either nominate them under an actually extant criteria (if it meets the requirements) or at RfD. Thryduulf (talk) 00:11, 9 December 2025 (UTC)[reply]

Should you be bold when tagging pages?

[edit]

I find a lot of drafts that are very clearly written with spammy language, but I feel fall just short of being blatant spam. See this as an example. Would it be best to avoid tagging these pages, or to tag them? ~2025-31416-56 (talk) 16:01, 9 December 2025 (UTC)[reply]

@~2025-31416-56 Should you be bold? Yes. The deleting admin should review the page for CSD eligibility, so in theory you can't get a page deleted incorrectly simply by tagging it. Do you need to tag that page? Meh. It's probably eligible for G11, or at least close enough that many admins would delete it. But tagging pages in draftspace is generally not a great use of time, since they will eventually be deleted by G13 anyways. In my view it's only helpful to tag drafts or userspace pages if they are actively harmful (see User:CoconutOctopus/How fast to speedy delete) or being repeatedly disruptively submitted to AfC. Toadspike [Talk] 09:32, 11 December 2025 (UTC)[reply]

Does T5 apply to Modules?

[edit]

Does WP:T5 also apply to unused Module subpages? I thought it did, but tried to tag one and got an error from {{db-t5}} that it should only be used in the Template namespace. Zackmann (Talk to me/What I been doing) 02:44, 11 December 2025 (UTC)[reply]

It does, but there is an error in the template code - see Template talk:Db-meta#db-t5. Thryduulf (talk) 04:17, 11 December 2025 (UTC)[reply]
Is it appropriate to amend the first statement of WP:T5 to say This applies to unused subpages of templates and modules, such as... (emphasis added for clarity here)? Does this require an RFC or broader consensus or can I BOLDLY make this change? Zackmann (Talk to me/What I been doing) 04:20, 11 December 2025 (UTC)[reply]
I think better would be a note at the start of the templates section that all the criteria apply to both templates and modules unless specified otherwise. This means we don't have to repeat notes if we get more T criteria. This sort of change doesn't require an RfC but it is best practice to mention it on the talk page (as you've done) and allow people chance to object/comment if they wish before making the change. Thryduulf (talk) 04:28, 11 December 2025 (UTC)[reply]
Sounds good. I'll give it a day then be bold. Can always be reverted. I'm not changing policy here (if I was I would feel I needed broader consensus) just trying to clarify what seems to be established policy already. Zackmann (Talk to me/What I been doing) 04:29, 11 December 2025 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Speedy_deletion&oldid=1326865261"