Wikipedia talk:Teahouse

Assistance for new editors

The Teahouse is frequently semi-protected, meaning the Teahouse pages cannot be edited by unregistered users (users with temporary accounts), as well as accounts that are not confirmed or autoconfirmed (accounts that are at least 4 days old with at least 10 edits on English Wikipedia).

However, you can still get direct assistance on your talk page. ; a volunteer will reply to you there shortly. Alternatively, you can contact an experienced editor by visiting your homepage and clicking "Ask your mentor a question about editing".

There are currently 0 user(s) asking for help via the {{Help me}} template:

Be mindful of LTA/CB885

Regular Teahouse answerers, please be mindful of the information at Wikipedia:Long-term abuse/ChronicleBooks885. Every time this user comes back I see people take the bait even though the signs are very obvious; be mindful and when this user appears, ignore them and hat their questions. Athanelar (talk) 20:00, 12 June 2026 (UTC)

@Athanelar What are the signs? How can you tell them from your typical newcomer? ^{In solidarity,} 🏳️‍🌈JohnLaurens333 (They/them • Ping me!) 05:44, 13 June 2026 (UTC)

@JohnLaurens333, the signs are described at the LTA page. You can also read through some of the past SPI cases to get a feel for them. —In solidarity with Wiki Workers United · ClaudineChionh (she/her · talk · email) 05:49, 13 June 2026 (UTC)

In brief, with registered accounts you're looking for a username like WordWordNumbers, WordWordWordNumbers, or occasionally just Wordwordword

With both registered accounts and TAs, their questions (both here and at the Teahouse) have a lot of recurring themes. There's the generic "I'm new, how can I contribute", the comparative questions "What is the difference between x and y" (usually asked at the refdesk, and usually the comparison is nonsense; like what's the difference between laptops and books) "Which is more reliable, x or y" (e.g., public libraries or Wikipedia), and questions relating to the difference between TAs and IPs and whether Wikimedia will ever go back to IPs.

As of late, they also have a habit of specifying that their questions should be answered "in your own words" (i.e., without linking them to project pages containing the answer)

They also generally never use less than 2 accounts at a time, sometimes up to 4 or 5. So if you suddenly see a string of this sort of question appear on the TH in quick succession from similarly-named accounts and a TA or two, that's a surefire tell. If you see them, hat their questions, keep an eye on the teahouse for a bit to hat any questions by other accounts, and report all the accounts to the SPI page for CB885. Checkusers normally find a couple of sleepers, and then sometimes they come back and do it again in a day or two, sometimes they'll be gone for a couple of weeks. Athanelar (talk) 06:14, 13 June 2026 (UTC)

What does "hat" their questions mean? SomeoneDreaming (talk) 16:08, 13 June 2026 (UTC)

@SomeoneDreaming See WP:Reference_desk/Science#Relations_between_Scientology_and_Science for a recent example. Basically, you WP:CLOSE the discussion started by a LTA rather than just revert their original edit. Mike Turnbull (talk) 16:46, 13 June 2026 (UTC)

Use {{Atop}} to close discussions and {{collapse top}} to, well, collapse them. Note that both have a separate template for the end of the content. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 11:40, 29 October 2026 (UTC)

@Athanelar Got it. I had already read the LTA page and was still unsure of how to tell them from newcomers, but this was very helpful, thank you. ^{In solidarity,} 🏳️‍🌈JohnLaurens333 (They/them • Ping me!) 16:18, 13 June 2026 (UTC)

I've just undone their latest effort. Mike Turnbull (talk) 10:23, 14 June 2026 (UTC)

Are they the same person who keeps asking "What if everyone on Wikipedia was a Checkuser?" or is that someone else? ^{In solidarity,} 🏳️‍🌈JohnLaurens333 (They/them • Ping me!) 15:34, 14 June 2026 (UTC)

That's them, yes. Athanelar (talk) 15:55, 14 June 2026 (UTC)

Lol I saw this immediately after taking the bait myself. Thanks for hatting. cyberdog958^Talk 03:52, 17 June 2026 (UTC)

Not sure how much effort goes into getting an llm to spin up a few paragraphs. CMD (talk) 04:42, 17 June 2026 (UTC)

I recommend Teahouse hosts who are unsure about whether they can spot this one apply for WP:TAIV. In solidarity, asilvering (talk) 05:18, 17 June 2026 (UTC)

Not eligible for TAIV yet, so checking my ability to spot them by hand: this one? how do administrators know that someone is vandalizing a page? M kuhner (talk) 18:16, 22 June 2026 (UTC)

Not one of their typical questions, the typing style has errors that are unusual for them, and there's no other accounts at the same time, nor any similar questions asked at other venues, so I'd say not them this time. Athanelar (talk) 08:38, 23 June 2026 (UTC)

I've just submitted a request since I spend a fair amount of time here, and watched one of their rampages unfold just after joining. I know I'm still fairly new even though I'm over the minimum requirements, so if admins don't feel comfortable granting it yet I understand. ChompyTheGogoat (talk) 05:39, 12 July 2026 (UTC)

Is this guy under the topic "new Wikipedia editer" CB885? Starlet! (Need to talk?) (Library) (Sandbox) 15:23, 23 September 2026 (UTC)

Doubtful. Their other edit has none of the tells. CB hasn't been active in a while, and it's totally normal for actual new editors to ask questions about being new. Try to look for the larger pattern, including how the requests are phrased. Multiple at once instead of just a one-off was also common, as Athanelar said above. ChompyTheGogoat [ Bleat | Munched ] 15:57, 23 September 2026 (UTC)

Good to know. Checked rf desk after, saw nothing. Probably just an actual newbie. Starlet! (Need to talk?) (Library) (Sandbox) 16:00, 23 September 2026 (UTC)

"Skip to bottom" anchor not working

I've noticed the above in the last couple of days, but have no idea whatever of how to fix it. Anyone? {The poster formerly known as 87.81.230.195} ~2026-43920-01 (talk) 10:22, 11 August 2026 (UTC)

The {{Skip to top and bottom}} template was removed from the Teahouse as it causes issues with the new parser, but this link referenced it. Replaced it with "#footer". Sjoerd de Bruin (talk) 10:34, 11 August 2026 (UTC)

Thank you very much, Sjoerd de Bruin. {The poster formerly known as 87.81.230.195} ~2026-43920-01 (talk) 13:49, 11 August 2026 (UTC)

Please remember to assume good faith r.e. AI-related questions

I know we all hate AI and treat any usage of it as an existential threat to the project. I mostly agree. But new editors don't know this, and the Teahouse is meant to be a place for new editors to get friendly Q&A from experienced users.

I think the editor at Wikipedia:Teahouse#Article Edited - request for independent review was treated fairly poorly by some hosts here. Wikipedia is facing an editor retention and readership crisis, let's try and be nice to the newbies even if they have fundamental misunderstandings of our policies and guidelines. qcne (talk) 10:24, 31 August 2026 (UTC)

Recent slew of "edit filter"-related questions

It looks like as at Sept 3, 2026, the Elden Ring wiki has a page that is redirecting their readers here to complain. —Tenryuu 🐲 ( 💬 • 📝 ) 04:25, 4 September 2026 (UTC)

Does anyone have a line to Fextralife's admins to let them know they've been vandalised? —Jéské Couriano v^_^v ^{Look out it's Jimothy!} 04:47, 4 September 2026 (UTC)

Looks like it's fixed now. —Tenryuu 🐲 ( 💬 • 📝 ) 19:29, 4 September 2026 (UTC)

See also Wikipedia:Administrators'_noticeboard/IncidentArchive1233#Off-wiki_UI_manipulation (the "I am a [sic] LTA and I was wrongly caught by the edit filter" thing), which is probably related. OutsideNormality (talk) 01:29, 5 September 2026 (UTC)

Extensive revdels of Teahouse

The edit history of the Teahouse shows extensive revdelling yesterday. (This includes a contribution of mine, which however is still there and readable; that's how I noticed it.) What happened? Maproom (talk) 07:53, 16 September 2026 (UTC)

One quick point -- do you see how there's two lines striking through all the edits? That means this isn't a standard revdel. Instead, these older revisions were oversighted. Oversighters remove material for a lot of reason; maybe somebody revealed a little bit too much about somebody (like a phone number or email address), but, for general safety/privacy reasons, the oversighters don't tell us what exactly they removed or why. GreenLipstickLesbian💌🧸 08:06, 16 September 2026 (UTC)

Yes, there's actually still a thread about that, paired with WP:CHILD me thinks. Lectonar (talk) 08:18, 16 September 2026 (UTC)

From what I can tell, someone added personally identifiable information to an article, which was oversighted, and then someone copied that information to the teahouse to protest the deletion of that information, however, there is much of the discussion that is oversighted, so I don't have a lot of info on this. BSH🌸 (talk) - (she/her) 12:23, 16 September 2026 (UTC)

The closed thread you're referring to is a separate matter. —Tenryuu 🐲 ( 💬 • 📝 ) 14:40, 16 September 2026 (UTC)

An editor made a comment that included a phone number from a person they received private correspondence from, but it went unnoticed for a while. —Tenryuu 🐲 ( 💬 • 📝 ) 14:39, 16 September 2026 (UTC)

Might as well take this opportunity to point to the PSA at Wikipedia talk:Teahouse/Archive 28 § Oversight. Rotideypoc41352 (talk · contribs) 15:39, 21 September 2026 (UTC)

"Assistance for new editors unable to post here"

I know it's not generally wiki's way to make sure every page looks perfect for every screen size etc, but every time I come to the Teahouse on my mobile, I see this post heading "Assistance for new editors unable to post here" which is displayed in a very narrow column splitting it up into two-letter chunks at a time, so it's like an almost-unreadable border on the left of the page. I think this is because there is no paragraph return after the box template, or for some similar reason meaning that the text wants to display beside the box instead of beneath it. I mention this in case there is some simple way to fix it. I assume it has not been apparent generally because most editors are on bigger screens. ~2026-42636-56 (talk) 07:16, 25 September 2026 (UTC)

This may be a WP:THEYCANTHEARYOU issue. (Don't let the shortcut fool you; that page discusses mobile bugs with respect to things like that.) —Jéské Couriano v^_^v ^{Look out it's Jimothy!} 07:31, 25 September 2026 (UTC)

Thanks for the suggestion but it didn't seem relevant — as far as I could tell it was just about not receiving communications. ~2026-42636-56 (talk) 22:01, 25 September 2026 (UTC)

I discovered that although I am unable to edit the page itself, there were three dots beside the problem heading allowing me to edit that section. I experimented by adding enough returns to shift the heading underneath the infobox, using three edits. I assumed I would be able to undo it afterwards if not successful. It worked in terms of allowing the heading to be readable because it brought it under the infobox instead of beside it, but left a large space visible under the box as well for some reason. I realised then that I couldn't revert (can't see how) or remove the spaces because they don't show up on subsequent edits. Apologies for this. I was too bold. Although maybe the space is acceptable since it does solve the unreadable text? ~2026-42636-56 (talk) 07:43, 26 September 2026 (UTC)

just to clarify, the returns I added before the text don't show up so I can't remove them and I don't know how to revert my edits. Also I couldn't explain my edits in edit history because the process was different to my usual experience of editing and I couldn't type into the comment box or review my edits in order to see the effect on the page. ~2026-42636-56 (talk) 07:49, 26 September 2026 (UTC)

I have restored the whitespace to the way it was before. I'm not sure what you're describing with being unable; I could do it just fine.

To revert your edits in future, go to the history tab of the page you're editing and click undo. crtbob@talk:~$ ls contribs 09:00, 26 September 2026 (UTC)

For clarity, I do see the issue you're describing on my phone: It's rendering the heading inline with the archive directory. I just tried fiddling with the display settings of the archive directory, but it didn't change anything. crtbob@talk:~$ ls contribs 09:11, 26 September 2026 (UTC)

@Crtbob and IP, did removing the padding, as I just did, help at all? Looks a bit naff on desktop but if that stops the display problem on mobile, desktop folks can get over it. In solidarity, asilvering (talk) 09:28, 26 September 2026 (UTC)

The problem is there again for me at present ~2026-42636-56 (talk) 09:48, 26 September 2026 (UTC)

Ok, reverted. I see I misunderstood you earlier, because on desktop (at least on "old" Vector) that archive box is so far away from that header I didn't even notice it. I don't have time to look into this further right now but presumably this is annoying mobile readers on every page with an archive box and has been for quite a while. In solidarity, asilvering (talk) 10:01, 26 September 2026 (UTC)

I've only noticed it on the Teahouse. ~2026-42636-56 (talk) 23:10, 26 September 2026 (UTC)

Still there :( crtbob@talk:~$ ls contribs 09:54, 26 September 2026 (UTC)

Thank you ~2026-42636-56 (talk) 09:49, 26 September 2026 (UTC)

Any other ideas for ways to fix this? ~2026-52276-33 (talk) 08:26, 5 October 2026 (UTC)

The Teahouse hasn't been protected for several weeks. I'm curious if any other regulars think we might be okay without that section for now? Perfect4th (talk) 18:40, 5 October 2026 (UTC)

I suppose we could comment it out for the time being. —Tenryuu 🐲 ( 💬 • 📝 ) 19:13, 5 October 2026 (UTC)

hello

how to ask questions and doubts ober here. i can't find the option pls help ~2026-52067-90 (talk) 14:36, 28 September 2026 (UTC)

You appear to have just asked a question and, in doing so, answered your own question. Unless I'm missing something? Cabayi (talk) 14:39, 28 September 2026 (UTC)

When you do have another question, please ask them at WP:TEA, not here as this talk page is for how the Teahouse operates itself. —Tenryuu 🐲 ( 💬 • 📝 ) 14:47, 28 September 2026 (UTC)

Professionals saying "I used ChatGPT"

I've seen professionals begin to mention here that they've used ChatGPT or similar services as if that were a good thing, and from this strange sort of forthrightness I'm beginning to suspect that it may (sometimes?) be better described as "written by a member of my staff, but cloaked as ChatGPT to avoid disclosing the conflict of interest". An intention to circumvent COI issues is the only thing I can come up with to explain their apparent eagerness to "wave the AI flag" up front in their very first contact here. (It's also unusual that they're calling it by its brand name, because most of those using it don't call it that; might be an extra-special bid to appear "authentic".) TooManyFingers (he/him · talk) 15:31, 11 October 2026 (UTC)

I don't see how this could help them on Wikipedia. I just think their behaviour seems odd if taken at face value. TooManyFingers (he/him · talk) 15:42, 11 October 2026 (UTC)

The I used AI mea culpas are interesting. It’s hard to know what to make of them. If you know it’s not allowed, why did you use it. Or, I’ve used it but I don’t want start again so maybe you’ll give me a pass. I wonder if it’s an AI recommendation. Mme Maigret (talk) 22:42, 29 October 2026 (UTC)

Wikipedia related printouts

I've found that c:File:Hannah Clover's Wikipedia printouts.pdf tends to be quite popular with people new to the concept of editing Wikipedia. I figured the Teahouse talk page was likely the best spot to spread the word if anyone wants to use it or repurpose it with tweaks for their own use. Clovermoss🍀 (talk) 00:05, 12 October 2026 (UTC)

To be deleted, maybe.

This should be deleted i guess. Krishnananda Dutta Chaudhury. No reliable citations.

The ref which is there has been written the family descendant only. EatingWoodenCupboards (talk) 10:00, 17 October 2026 (UTC)

Hi EatingWoodenCupboards. For reference, this is the talk page for the Teahouse; it's not the Teahouse itself. This talk page is for discussing things (e.g., proposing improvements) related to the Teahouse. This page isn't really the place for posting questions about things other than the Teahouse. If you feel the article you're asking about meets the criteria for WP:DELETION, you can nominate it for deletion using Wikipedia:Articles for deletion or tag/propose it for deletion using one of the other deletion process. If you're not sure, try asking for input at Wikipedia talk:WikiProject India. -- Marchjuly (talk) 14:01, 17 October 2026 (UTC)

User:Akuru Kevin

Could a host please invite Akuru Kevin to the Teahouse? They seem to be editing in good faith but most of their edits ended up being non-constructive. I reverted the obvious ones, but several more might need a more careful review. Thank you. NguoiDungKhongDinhDanh 14:42, 21 October 2026 (UTC)

 Done BlueStaticHorse (talk🌸) 15:12, 21 October 2026 (UTC)

I have a problem with Sinking of the Titanic

Moved to Wikipedia:Teahouse § I have a problem with Sinking of the Titanic

Archiving

I just wanted to note I have done a round of archiving due to the bot outage. Depending on how long the bot outage lasts for, more archiving may need to be done. 45dogs (they/them) (talk page) (contributions) 23:10, 22 October 2026 (UTC)

Thank you! If it lasts longer, I can do this as well, if needed. BlueStaticHorse (talk🌸) 23:19, 22 October 2026 (UTC)

Teahouse invite (suggestion)

We get a lot of the same questions here in the Teahouse. But there is one in particular, I'm wondering if we could address in the Teahouse invite. When an AfC article get declined, the banner invites them to the Teahouse but I'm wondering if one thing that should be added to that notice is not to ask the Teahouse for a pre-review. (There may be a better place for this suggestion, like maybe the AfC talk page or the village pump.) Mme Maigret (talk) 01:18, 27 October 2026 (UTC)

An idea that certainly merits discussion, Mme Maigret; but better raised and discussed on Wikipedia talk:Teahouse. Perhaps move it there? -- Hoary (talk)^WWU 👍︎ 02:16, 27 October 2026 (UTC)

Here's the template if anyone wants to look at it: Wikipedia:Teahouse/AfC Invitation. qcne (talk) 10:41, 27 October 2026 (UTC)

Currently:

For peer support regarding all other editing, sourcing, or policies, visit the Teahouse, a friendly question and answer hub for new editors. We'd love to see you there!

A first attempt:

For advice about other matters, visit the Teahouse, a friendly question-and-answer hub for new editors. NB The Teahouse doesn't do "pre-reviews": it won't give your draft a general look-over before you submit it (using the Resubmit button).

Bearing in mind the hostility shown to alleged uses of an LLM when asking questions or for help, a second attempt:

For advice about other matters, visit the Teahouse, a friendly question-and-answer hub for new editors. NB (i) The Teahouse doesn't do "pre-reviews": it won't give your draft a general look-over before you submit it (using the Resubmit button). (ii) Please don't use an LLM ("AI").

No doubt you, Mmemaigret, or some other editor(s) can improve on one or other of these. -- Hoary (talk)^WWU 👍︎ 04:05, 27 October 2026 (UTC)

I'd certainly support battering people over the head as much as possible with messaging about not doing both of these things. Not sure how much good it'd be because AI-powered COI editors don't tend to pay all that much attention to the things we try to tell them in my experience, but it can't hurt to try. The number of identically-worded AI generated review requests on the Teahouse in the past few days has been completely ridiculous. I counted 3 in one day. Athanelar (talk) 08:03, 27 October 2026 (UTC)

Athanelar, bogus, LLM-generated pseudo-references and the like are one thing, LLM-assisted review or other requests are to my mind quite another. I don't care much about the latter (I'm unfazed by such minor phenomena as the use of asterisks to indicate bold-toggling), and they do have the plus of intelligibility. Well, I do care that they are review requests, betraying a complete lack of effort to understand what's required; however, if they're identically worded, so much the better: their uniformity should make them slightly easier to digest and dismiss. -- Hoary (talk)^WWU 👍︎ 10:29, 27 October 2026 (UTC)

@Hoary Can I suggest the following change:

For advice about other matters, visit the Teahouse, a friendly question-and-answer hub for new editors.

NB

(i) The Teahouse doesn't do "pre-reviews": we won't give your draft a look-over before you re-submit it

(ii) AI use is strictly prohibited on Wikipedia to write articles or Teahouse comments.

Mme Maigret (talk) 09:36, 27 October 2026 (UTC)

Mme Maigret, your (ii) sounds slightly unidiomatic to me. How about "Using AI to write articles or Teahouse comments is strictly prohibited"? However ... if I were a NooB, then reading that, I might think "Ah, so drafts are OK then. I'll use an LLM to create my draft, submit it, and after it's accepted I'll use the article's talk page for my suggestions for improvements. Sort of like the COI matter." Interestingly (or invitingly), the page WP:Writing articles with large language models nowhere mentions "draft". As I've written above, I'm not sure that LLM use is as toxic as others make it out to be. But if it is, shouldn't it be prohibited for drafts as well? (I imagine that this would have been discussed in WT:Writing articles with large language models, but it's been a long day and I'm disinclined to search.) -- Hoary (talk)^WWU 👍︎ 10:55, 27 October 2026 (UTC)

@Hoary Happy to change it to "Using AI to draft/write articles or comments is strictly prohibited". Mme Maigret (talk) 11:48, 27 October 2026 (UTC)

Mme Maigret, I'm not sure that using AI to draft articles is prohibited; and if it isn't already prohibited, then a quick decision here to prohibit it probably wouldn't suffice. As I've said, WP:Writing articles with large language models doesn't mention drafts. (In an earlier message here [10:55], I may have suggested that later I'd take a more serious look into its megabytes of "talk". But now that I've had a good night's sleep, I find that I can't be bothered. Sorry! Athanelar, do you know about the position, if any, on LLM use for drafts?) -- Hoary (talk)^WWU 👍︎ 00:16, 28 October 2026 (UTC)

@Hoary LLM is one of the grounds for declining an article at AfC. The AfC reviewing instructions say "The use of large language models to generate or rewrite article content is generally prohibited under the LLM guideline, including in draftspace." Mme Maigret (talk) 00:20, 28 October 2026 (UTC)

I recall there was a lengthy discussion about it somewhere, though fucked if I remember where. The conclusion was that because prose in draft articles is explicitly intended to become article prose, NOLLM applies. Userspace is more of a grey area, but draftspace and userspace drafts are a no-go. Plus indeed as Maigret has said, LLM use is a (very often used) reason for decline; see "ai" under Template:AfC submission/comments. Athanelar (talk) 00:38, 28 October 2026 (UTC)

Sounds reasonable. Let's consider for a moment the implications of expressly prohibiting LLM use for articles but expressly permitting it for drafts. That would make the draft reviewer responsible for the de-LLM-ification of any draft they thought potentially worthy of acceptance. Hmmf! I for one would resign from draft reviewing pronto. -- Hoary (talk)^WWU 👍︎ 01:03, 28 October 2026 (UTC)

My preferred solution to this:

1. Draft reviewers really should be the ones responsible for de-LLM-ification.

2. Draft reviewers' de-LLM-ification should consist only in removing potentially LLM-influenced parts, and handing back the (potentially empty) draft to the submitter. TooManyFingers (he/him · talk) 18:55, 29 October 2026 (UTC)

@Mmemaigret I recommend not using the abbreviation "NB". I know what it means, but I suspect that lots of "kids these days" will not understand it. You could use "by the way", or you could just remove "NB" itself (and don't use anything in its place). Understanding "NB" might also be country-specific.

Your wording after "NB" is, of course, important. David10244 (talk) 08:56, 31 October 2026 (UTC)

@Hoary 👀 Mme Maigret (talk) 09:53, 31 October 2026 (UTC)

@David10244 The kids will just have to learn Latin. Mme Maigret (talk) 09:55, 31 October 2026 (UTC)

David10244, "Note that". -- Hoary (talk)^WWU 👍︎ 10:13, 31 October 2026 (UTC)

Constant AI use by editors with conflict of interest

In recent months, multiple of the questions written here are people with conflict of interest wanting an article (which they are not notable) use AI to generate their questions (and are collapsed by other editors) while not admitting it. The repetitiveness is getting repulsive and some other editors especially don't like the AI use. (while suitable AI use is great and may even be efficient, these editors' AI use is not okay)

Is there anything we can do about this (like adding some kind of banner or other more systematic solutions)?

I am not sure if this issue has been raised before, and if it was, please direct me to the relevant discussion page.

Sincerely, 16dvnk (talk) 13:26, 27 October 2026 (UTC)

P.S. It seems that the above discussion has partially addressed the issue. Oh well. 16dvnk (talk) 13:28, 27 October 2026 (UTC)

Aside from the AfC invite banner change being discussed above, might it be worth adding an AI usage reminder to to the page notice, which is shown when editing Wikipedia:Teahouse? Something like "Do not use an LLM to generate messages here, it's against the rules and we will get upset at you" (more formally, of course). crtbob@talk:~$ ls contribs 22:27, 27 October 2026 (UTC)

A message in bold like "Messages generated by AI will be collapsed and not answered" would work, I think. —Jéské Couriano v^_^v ^{Thriller Night!} 22:42, 27 October 2026 (UTC)

A bold message doesn't fit the aesthetic of the Teahouse. However, I boldly added your message to Wikipedia:Teahouse/Host lounge/Announcements/1. Apparently the template at the top of the Teahouse is a series of templates almost like Russian nesting dolls. Mikeycdiamond (talk) 23:40, 27 October 2026 (UTC)

will be collapsed and not answered

But will they, though? Most of the obviously AI-generated requests from the past few days have been answered along with a slap-on-the-wrist "Don't use AI to talk to us please" warning. crtbob@talk:~$ ls contribs 01:59, 28 October 2026 (UTC)

I feel like if the request seems to be otherwise made in good faith it's nice to answer them. It does get tiring though. I worry sometimes that I'm subtly encouraging bad behavior when I do that, but also I don't like denying a request I could answer on a technicality. Packrat3860 (talk) 02:17, 28 October 2026 (UTC)

My comment wasn't necessarily a statement for or against answering them; I am currently only pointing out the inaccuracy of the message. It currently reads that AI content will be collapsed and may not be answered, but that's not currently the case. I feel we need to either change the message to align with our approach to those questions, or change our approach to those questions to align with the message. crtbob@talk:~$ ls contribs 05:45, 28 October 2026 (UTC)

@Jéské Couriano I think this message suggests go ahead and try it on and then we'll collapse it and we might still answer it. I think it would be better to just say no AI use and then have a policy of collapsing once detected. (Having said that, I don't personally have an issue with people using AI in comments particularly if it helps them express a point or helps with a language barrier. What really gets me are the people who can't seem to stop after they've been asked not to use it or who lie about it when it's obvious they are using.) Mme Maigret (talk) 08:50, 28 October 2026 (UTC)

My acceptance of AI use is similar to what @Mmemaigret has said on the above. I edited the message to say "may not be" because currently, some people still answer it. 16dvnk (talk) 10:34, 28 October 2026 (UTC)

Another concern: would the target audience know what "collapsed" means in this context? Is there an explanatory wikilink that can be used here? crtbob@talk:~$ ls contribs 16:03, 29 October 2026 (UTC)

"hidden"? Or "removed"? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 17:48, 29 October 2026 (UTC)

I do not agree that we should collapse even AI massages that are first posts (except in cases of egregious length; even then only the excessive part should be collapsed).

The Teahouse is the one page where we are supposed to welcome people who don't yet know about our rules and culture.

If we advise someone not to use AI and they continue to do so, then by all means collapse their messages, close the section, or, after a further warning, block them. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 11:34, 29 October 2026 (UTC)

I agree with Andy. The Teahouse should be somewhere where "one strike and you're out" should not apply. Our first duty as hosts is to answer the question asked, assuming we can understand it. By all means politely point out the issues with AI if appropriate but hosts should be aware that WP:NOLLM is guidance, not policy. Mike Turnbull (talk) 12:38, 29 October 2026 (UTC)

How do you suggest we word it to discourage AI use in the Teahouse? Mikeycdiamond (talk) 15:22, 29 October 2026 (UTC)

The section above has some suggestions. However, no matter what wording we use, a large number of newcomers won't have read anything before they go ahead and post here or at the Help Desk. You can't run any help venue on the basis of RTFM. Mike Turnbull (talk) 16:52, 29 October 2026 (UTC)

"One LLM strike and you're out" would be a terrible plan, I agree. But "two more LLM strikes after being warned is a compulsory indef" would be perfectly reasonable IMO. TooManyFingers (he/him · talk) 18:22, 29 October 2026 (UTC)

Using the Teahouse to discourage/prohibit LLM use in drafts and articles: Yes, good. But why discourage/prohibit LLM use in the Teahouse? However wrong-headed they might be, (seemingly) LLM-generated questions and requests (Q/R) are at least comprehensible. Even after putting aside Q/R that look like mere trolling, I can't say the same for many (seemingly) not-LLM-generated Q/R. -- Hoary (talk)^WWU 👍︎ 22:42, 29 October 2026 (UTC)

Because people who use AI to ask a question don't understand what they themselves have posted, let alone the answers they might get. They are not participating as people anymore, but pigeons pecking at the experiment button to get the food reward. I'm inclined to start answering AI-worded questions with plausible-but-very-wrong answers. TooManyFingers (he/him · talk) 01:35, 30 October 2026 (UTC)

Some of the people who use AI to ask a question don't understand what they themselves have posted, let alone the answers they might get. -- Hoary (talk)^WWU 👍︎ 10:23, 31 October 2026 (UTC)

Comment Sometimes, genuine (albeit very little) good faith editors may use AI to translate things from their language . That is a right way to use AI, but the current message does not address that.

Other than proper use, things where the author does not even understand the AI and refuses to admit it should not be allowed. 16dvnk (talk) 13:23, 31 October 2026 (UTC)

Fundamentally, AI is a good tool. But only if you use it correctly and ethically. Unfortunately, a recent and significant batch of these editors here are not disclosing, and are making excuses after excuses after excuses and denying after being caught. In the end, they continue to disrupt Wikipedia. Some do not even read what the AI is saying, and they lack too much experience in order to prompt them correctly. Good faith uses should be accepted, but I am seeing far too much bad faith uses and no intention to correct themselves after being caught. 16dvnk (talk) 13:30, 31 October 2026 (UTC)

Hoary - you're right that it's only some of them. The others are lazy and don't give a shit, and I'm not too sure whether either is better than the other. TooManyFingers (he/him · talk) 19:54, 31 October 2026 (UTC)

Many people won't make the distinction between using AI in the Teahouse and the rest of Wikipedia. If we want the use of AI on Wikipedia to end, we must make a clear stance that it is unacceptable. I get, and now agree with, the argument that we shouldn't automatically collapse AI questions when people haven't been warned/told about Wikipedia's policies on AI, but we must take a clear stance on AI. If someone can't ask a coherent question without AI, perhaps competence is required might apply. Mikeycdiamond (talk) 01:49, 30 October 2026 (UTC)

I agree. The hushed and fearful tone surrounding CIR comes from a bygone era, when a person who didn't know any English would never have thought of coming here. If our virtual forebears roll in their virtual graves because we now need to hand out CIRs and bans like they were candy, so be it. TooManyFingers (he/him · talk) 03:48, 30 October 2026 (UTC)

"a bygone era, when a person who didn't know any English would never have thought of coming here"—Having ben here since October 2003, and still very much alive, when exactly was that, I wonder? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 14:48, 31 October 2026 (UTC)

I hope my idea was better than my words. There is a difference. There's a new group of people who haven't reached the written English proficiency expected of English-speaking four-to-six-year-olds. Twenty years ago, those people would not have tried to edit Wikipedia so often or in such numbers. TooManyFingers (he/him · talk) 19:49, 31 October 2026 (UTC)

Faq this

Requesting a notability assessment for a possible biography

Their alleged LLM use aside, such requests are tiresome -- but of course they merit an informative, helpful and courteous answer. And they're frequently asked.

So I recommend their elevation to FAQ status.

As first try -- greatly indebted to Athanelar, Wikipedia:Teahouse#c-Athanelar-20261028221100-Prcohen01-20261028220100 -- I give you:

"Notability assessments" are not provided at the Teahouse. To the contrary: you are expected to (i) familiarise yourself with the relevant notability criterion (start by reading and digesting "Wikipedia:Notability", continue by reading and digesting the seemingly relevant pages to which this links) and then (ii) have your draft demonstrate that the subject of your proposed article meets this criterion.

(The lack of mention within that of either LLMs or COI is very deliberate.) Comments? -- Hoary (talk)^WWU 👍︎ 00:55, 29 October 2026 (UTC)

I think it's somewhat impossible to separate LLM use from this question—it is only asked so frequently as of late because chatbots are telling people to ask it—but that academic point aside I would support making an FAQ about it (and I'm flattered that you'd base it on my wording!) I'll probably write up a nice boilerplate in my userspace at some point too for easy use. Athanelar (talk) 01:17, 29 October 2026 (UTC)

The main reason why I separate it from LLM is that definitive proof of LLM use is hard to show in a first response, while mere accusation of LLM use makes it very likely that what was intended to spur the questioner/requester to some hard thinking about "notability" will instead spur them to a denial (true or false) of LLM use or an argument for it. (But I have to admit that a secondary reason is my appreciation of LLM-assisted questions/requests: unlike many of their 100% human-powered alternatives, they're easy to understand.) -- Hoary (talk)^WWU 👍︎ 02:02, 29 October 2026 (UTC)

I also noticed that comment of yours, @Athanelar, and mentally bookmarked it to draw on if necessary! I would support adding this to the FAQs as well. Suðurhafsljósæta (talk) 10:25, 29 October 2026 (UTC)

I suggest a link to WP:42 as well as, or preferably instead of, WP:N. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 11:25, 29 October 2026 (UTC)

Not "instead of", please, Andy Mabbett. Consider how "42" starts. Stripped of typographic exuberance, it's: Articles generally require significant coverage in multiple, reliable sources that are independent of the subject. I expect that many readers will react with: "Ah, generally. So not always." The more dutiful of readers will keep reading, whereupon they'll thrice encounter "usually" but won't see any mention of what kinds of exception can be acceptable. (Meanwhile, if WP:N's verbosity makes you unenthusiastic about it, that makes two of us.) -- Hoary (talk)^WWU 👍︎ 23:13, 29 October 2026 (UTC)

You can, of course, propose or make a change to:42. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 23:17, 29 October 2026 (UTC)

My suggestion may be counterproductive, too difficult to do, or all sorts of other problems. I'm happy for it to be booted out the door for any such reason.

A prominent link near the top of the Teahouse that leads to a separate page - the Biographies Teahouse. On there, all of the biography-specific stuff with which we avoid cluttering the Teahouse could be on full display.

A biography written by a newcomer necessarily forces them to jump through a lot of hoops. In other contexts, taking the newbies and "showing them the ropes" is normal. Here, we'd be "showing them the hoops" from the beginning. The way it's done now, each new hoop is a rude surprise they get from some individual, requiring them to perform a new round of disbelief, arguing, and blaming the messenger. TooManyFingers (he/him · talk) 18:40, 29 October 2026 (UTC)

From what I've seen, most of the biography related questions over the last few days have been either:

1. Why did my biography get rejected?
2. Is this person notable?

#1 should probably be redirected to WP:AFCHD, to lessen burden on Teahouse answerers, and because the people over there are probably better equipped to answer those questions (side-ish note: should there maybe also be a note in the teahouse header or editnotice to take rejected draft questions there instead?)

AFAIK, #2 doesn't have a specialised place, which I could certainly see becoming a thing. Imagine a beginner-oriented place where beginners can seek advice on whether their idea for a article is notable – they explain what it is, provide their sources, and then more experienced editors can advise whether or not it looks notable. Basically just a new place for the questions that have been filling up the Teahouse for the past few days.

I do see how this could be a little excessive, though. The existence of yet another help forum could be confusing to newcomers, and the Teahouse is still probably a fine place to have notability assessments.

_{IDK though, i'm really just writing stuff because TooManyFingers's idea was interesting and I was building upon it, no idea if this is feasible} crtbob@talk:~$ ls contribs 19:29, 29 October 2026 (UTC)

There is little point redirecting people to WP:AFCHD, if AfC templates are redirecting them here. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits; 19:38, 29 October 2026 (UTC)

You're right, I'm sure, that those are the questions. But one of the biggest problems in this topic is people cluelessly asking the wrong question and needing to be answered with something else entirely. TooManyFingers (he/him · talk) 20:04, 29 October 2026 (UTC)

The actually-relevant response to "Why did my biography get rejected?" - if we're assuming they did ask the right question - is the single sentence "They already told you why", followed by closing the topic to further responses. TooManyFingers (he/him · talk) 20:36, 29 October 2026 (UTC)

TBH, the folks who merely ask if someone is notable feel fine to me. Wikipedia has a finicky concept of notability (for good reason) and it's a natural thing to help newer folks with. It's the folks who want to try to debate the Teahouse into recognizing their notability (usually with obvious LLM assistance) that have me rolling my eyes. Packrat3860 (talk) 23:30, 29 October 2026 (UTC)

The problem is that these people usually haven't even attempted to understand it, they just come and ask for a "notability assessment" and it's like, no; assessing notability is literally the whole difficult part of writing an article. Sure, we can give people some pointers on how to carry out that assessment themselves, but it is an article author's job to assess and demonstrate notability, not to have somebody else do that part for them and then they just write the prose. LLMs are encouraging people to think that "notability assessment" is some kind of wizardry that only us Wikipedians can do, and that they as laypeople need to have us do that part for them. Athanelar (talk) 23:58, 29 October 2026 (UTC)

@Crtbob In my personal opinion, the answers at AFCHD are often quite abrupt, bordering on rude. The answers at the Teahouse are generally more polite. David10244 (talk) 09:04, 31 October 2026 (UTC)

Or: Mandatory use of an as-yet-unwritten Biography Wizard that uses "evil" modal dialogs to force Step 1 to be completed before Step 2 can be viewed, and Step 1 is "Submit your sources for review and get them approved". TooManyFingers (he/him · talk) 19:06, 29 October 2026 (UTC)

For the record, I just took a quick count and there are currently 14 AI-generated notability questions active on the Teahouse. The scale here is getting ridiculous. Athanelar (talk) 09:31, 30 October 2026 (UTC)

Chalk up one more this morning at least.

As far as scale goes I think a big part of the issue with trying to be "lenient" with LLM use is that if it's allowed to quickly drowns out everything else. I don't where to draw the line because, as discussed above, the Teahouse is meant to be welcoming for folks who are genuinely just not aware of en-wiki's policies and guidelines so brusquely rebuffing them for violating that policy cuts against the mission. But if the Teahouse becomes a place where someone will review your slop draft coi biography then that will eventually become the only thing it's for. Packrat3860 (talk) 17:00, 30 October 2026 (UTC)