Wikipedia:AI noticeboard/Archive 8

 Umar2z (talk · contribs · logs · block log) (28 April 2026)
Editor has been blocked from editing the mainspace (articles only) for AI/LLM violations. Due to lack of communication. Dr vulpes (Talk) 00:36, 28 April 2026 (UTC)

Umar2z (talk · contribs · deleted contribs · logs · filter log · block user · block log)

Hundreds of most likely AI edits since 2023 - this appears to be the first one -- and continuing to this day. Note also that this user has edits prior to 2023, which read quite differently in writing style. Gnomingstuff (talk) 04:05, 27 April 2026 (UTC)

Blocking from mainspace until they can explain their behavior. Dr vulpes (Talk) 05:33, 27 April 2026 (UTC)

Thank you for raising this and i really appreciate the gesture because it made me went back to read and understand the WP:Writing articles with large language models - (WP:LLM). I wish to assure you that my intentions are good, just for contributing my quota to open knowledge. Note from this which goes back as 2023, my editing skills have increased overtime away from the use of LLM, thanks to constant training and retraining. I am also open to more guidance on this.

Also prior to this, i have never received any warning regarding the use of AI in my edits, I just checked my mail now only to see the message notifying that i have being blocked from the mainspace.@Dr vulpes: Umar2z (talk) 20:31, 28 April 2026 (UTC)

Good day @Dr vulpes:, i am yet to get any response from you with regards to my block from editing the main space, this is the third month. I will be glad to hear a feedback from you. Thanks

 Zeptentrion (talk · contribs · logs · block log) (28 April 2026)
Editor has been blocked from editing the mainspace (articles only) for AI/LLM violations. blocked from mainspace Dr vulpes (Talk) 20:04, 28 April 2026 (UTC)

User:Zeptentrion has created multiple articles using what looks like AI (1, 2, 3, etc.), then repeatedly used AI on article talk pages (see Talk:Amaël L'Étang, Talk:Chibuzo Agbo, etc.) to gish-gallop and deflect from the issue.

They requested a third and fourth opinion, both of which agreed with my position, before requesting a fifth opinion. Zeptentrion says they have not made any prohibited use of LLMs under the current guideline, not once but twice (1, 2), but I find this highly dubious. Wikipedia:Do not lie about AI seems appropriate here. This doesn't seem like an editor interested in our project. JTtheOG (talk) 19:48, 28 April 2026 (UTC)

Well, we're waiting on the seemingly-inevitable consensus about WP:LLMPROD, but for the time being I'm going ahead and draftifying all of this editor's articles based on their pblock. Athanelar (talk) 01:11, 30 April 2026 (UTC)

 Done

I've draftified all of this editor's article creations, and reverted as many of their large additions as I could be bothered to manually crosscheck and excise from their articles (i.e., not many; the problem is that so many of these sports player articles are constantly updated with the latest stats by fans that the big AI additions get buried, and then it's a choice of either mass-reverting all the stat updates, or manually checking what text was added and cutting it out, which isn't quick or easy.) Athanelar (talk) 01:23, 30 April 2026 (UTC)

 Itsdanielsultan (talk · contribs · logs · block log) (8 May 2026)
Articles moved to draftspace pending review and cleanup. 

 Cleanup has been requested.

Itsdanielsultan (talk · contribs · deleted contribs · logs · filter log · block user · block log)

Itsdanielsultan has been creating WP:AISIGNS-heavy articles. Previous talk page notice went unanswered. Some fairly egregious examples just from their last three edited articles:

Draft:Ron Lenyk Springfield Park: No independent assessments of the park's design, maintenance quality, or environmental condition were identified in published reliable sources at the time of writing. Some user reviews aggregated on third-party platforms noted occasional wasp problems during warmer months.

Tunnel Vision Challenge: The contest drew attention on social media in January 2026 but received limited coverage from major news outlets at the time of its announcement., cited to , which of course mentions neither social media nor news outlets.

Dosti Aisa Naata: The song remains part of the cultural memory of Pakistanis who grew up watching PTV in the 1970s and 1980s. Its continued performance at events such as the Children's Literature Festival in 2016 and the ARY News Independence Day broadcast in 2018 reflects its status as a nostalgic touchstone for that generation.

~ A412 ^talk! 17:25, 2 May 2026 (UTC)

WP:ANI#AI-generated articles by Itsdanielsultan Kowal2701 (talk, contribs) 17:32, 2 May 2026 (UTC)

yep also note the AI edit summaries Gnomingstuff (talk) 19:56, 2 May 2026 (UTC)

See Gold market in India. It's pretty obvious that it's AI... even the user's user page seems to have been AI generated (although I don't think that's not allowed). Now, the major problem is, is that the user has created the same article in seven other Wikipedia's. I plan to nominate the English Wikipedia article for speedy deletion, however, what do we do with the rest? Do we notify administrators on the other wikis, but at the same time, every Wikipedia has its own policy, so this just makes the situation even more complicated. Fortek67 (talk) 21:53, 29 May 2026 (UTC)

Ivolskiy also seems to be responding to me with AI generated comments. See User talk:Ivolskiy. I have also now nominated the article for a speedy deletion. Fortek67 (talk) 22:27, 29 May 2026 (UTC)

When you nominate as db-g15, you must include a reason with your evidence. Otherwise an administrator has no choice but to decline the nomination. It isn't obvious to me that it's LLM generated but I'll leave your db-g15 tag there to give you a chance to fix it. Once the tag is removed, it cannot be put back. ~Anachronist (who / me) (talk) 03:03, 30 May 2026 (UTC)

I've filled in the reason parameter on Gold market in India. It's the first time I opened that article and just wanted it to be filled in before someone declines the tag, @Fortek67 should add anything I missed. You should normally give as detailed reasons as possible in these, basically make it as easy as possible for the patrolling admin to handle the request. Only mention things that are valid G15 criteria, anything else should go in an {{AI-generated}} tag instead. --gurkubondinn 06:21, 30 May 2026 (UTC)

Thanks for the help, gurkubondinn! Fortek67 (talk) 11:18, 30 May 2026 (UTC)

Sorry about that, will add a reason next time. Fortek67 (talk) 11:18, 30 May 2026 (UTC)

Probably, the eswiki article can be speedy deleted as well. Their speedy deletion policy includes an LLM rule under speedy deletion criterion G12. –LaundryPizza03 (dc̄) 22:29, 29 May 2026 (UTC)

Well, thanks for telling me about that! I have nominated the Spanish article for a speedy deletion. I don't speak Spanish though so hopefully that won't create any problems. Fortek67 (talk) 22:37, 29 May 2026 (UTC)

For future reference, it might be a better idea to reach out to the local WP:EMBASSY of a foreign Wikipedia when you don't speak the language. Just think about how much trouble we deal with from editors who can't speak English. Not saying that you've done anything wrong or something, just a friendly suggestion. --gurkubondinn 22:50, 29 May 2026 (UTC)

Oh wow, this is a very useful tool, will report to them. Thank you so much! Fortek67 (talk) 22:51, 29 May 2026 (UTC)

 Louloureagan9 (talk · contribs · logs · block log) (4 May 2026)
Articles moved to draftspace pending review and cleanup. 

User:Louloureagan9 is creating LLM-generated articles on malls at a rapid pace and is not responding to multiple talk page messages. Boise Factory Outlets and Portage la Prairie shopping Centre are some examples. Clear WP:OVERATTRIBUTION first of all... JTtheOG (talk) 15:57, 3 May 2026 (UTC)

 QINGSHAN XUE (talk · contribs · logs · block log) (2 May 2026)
This issue is currently under review. Started the list (present-November 2025) and reverted a couple edits. I₂Overcome ^talk 08:37, 2 May 2026 (UTC)
Added October and a few clear cases from earlier. I₂Overcome ^talk 19:22, 3 May 2026 (UTC)

Articles requiring cleanup and evaluation

CleanInternational Steels Limited  I2Overcome

CleanPrecision Engineering Complex  InfernoHues  Already cleaned by PeerBaba

CleanFirst Women Bank  I2Overcome

CleanAisha Steel  I2Overcome

CleanNational Logistics Corporation  I2Overcome

Needs reviewPakistan Digital Authority

StubifiedSindh Safe Cities Authority  I2Overcome

Needs reviewNational Clearing Company of Pakistan

DeletedMain University Road

Needs reviewDraft:Mega Motor Company

CleanMashreq (bank)  I2Overcome

CleanLucky Motor  I2Overcome

Needs reviewDraft:Sofitel Le Scribe Paris Opéra

Needs reviewDraft:Organic Meat Company

CleanKamran Khan (journalist)  I2Overcome

CleanPakistan International Airlines  Chipmunkdavis

CleanARY Digital Network  I2Overcome

Needs reviewDraft:PIA Holding Company Limited

CleanAHS Tower  Chipmunkdavis

Needs reviewDraft:Excise, Taxation and Narcotics Control Department

CleanHub District  Chipmunkdavis

CleanBagh Ibne Qasim  I2Overcome

CleanBoat Basin Park  I2Overcome

AfDBeach Park, Karachi  InfernoHues  nom'd for LLMPROD

CleanLotte Chemical  I2Overcome

CleanLotte Chemical Pakistan  I2Overcome

CleanMahan Air  Dr vulpes  cleaned up by GeminiHeron. Dr vulpes (Talk) 01:33, 3 May 2026 (UTC)

CleanPort Qasim Power Project  I2Overcome

The usual, a lot of AI-generated edits and new articles. They seem to have started in mid-2025, the stuff earlier than that doesn't especially seem like AI, haven't checked too deeply though. Gnomingstuff (talk) 07:23, 2 May 2026 (UTC)

 AnnaBSP (talk · contribs · logs · block log) (8 May 2026)
This issue is currently under review. 

In progressCybersecurity engineering  Dreamyshade

Needs reviewRobotics engineering

Needs reviewComputer network engineering

Needs reviewArtificial intelligence engineering

AnnaBSP (talk · contribs · deleted contribs · logs · filter log · block user · block log)

This person created four long articles in October-November 2024 as a brand-new user, each in one fell swoop, with language patterns that sound like a LLM was involved. Other people have put some work into these articles over the past couple years though, and the articles seem quite salvageable - for example, I haven't found hallucinated citations, just some low-quality citations that a new editor could easily have not realized were low-quality, such as Forbes articles and predatory journals. I believe this was a good-faith effort; I can't find any obvious spam angle. So I'm asking here for help with cleanup.

• Cybersecurity engineering - I'm working a bit on this one and will probably propose merging it into security engineering.
• Robotics engineering
• Computer network engineering
• Artificial intelligence engineering

Dreamyshade (talk) 00:41, 5 May 2026 (UTC)

 Cleanup has been requested.

https://en.wikipedia.org/wiki/Song_of_the_Flame_(film)

The plot summery portion for this 1930 film seems quite extensive, even for portions that have been completely lost for decades.

"After the daring challenge, Count Boris—the Prince’s aide—intervenes: “Who dares put our Prince’s valor to the test?” The strange creature—who is none other than Anuita, the Flame—cries out: “I do!” “And what if the Prince proves his valor?” “I shall give him a kiss.” The Prince prevails; then, sweeping her into his strong arms, he lifts her onto the saddle of his horse and gallops away to claim the promised kiss far from that place. In that moment, their attraction deepens into genuine love. " (Note: None of this is in the surviving audio of the film)

What I'd like to think is happening here, is someone fed an AI the script of the movie, but after an hour of searching I can't find the screenplay, so I am doubtful.

Could it really be some obscure summery from the early 1900s, or is it just AI hallucination? I genuinely do not know.

Incredibly sorry if this is not the correct place to post this, I am not entirely sure. — Preceding unsigned comment added by Glarph (talk • contribs) 02:03, 6 May 2026 (UTC)

Whatever this is, it's excessive and the plot should be shortened. I agree that it's likely AI somehow but can't really say for sure. Lijil (talk) 05:41, 6 May 2026 (UTC)

The Curly quotes suggest it was at the very least copied from somewhere else. CMD (talk) 06:34, 6 May 2026 (UTC)

These large plot expansions go back to July 2025. CMD (talk) 06:36, 6 May 2026 (UTC)

mobile does curly quotes (though the edits don't have the mobile tag) Kowal2701 (talk, contribs) 13:18, 6 May 2026 (UTC)

ColorProcess (talk · contribs) Kowal2701 (talk, contribs) 13:20, 6 May 2026 (UTC)

Was previously asked about this (in general) last August and said it was their own words. ClaudineChionh (she/her · talk · email · global) 04:24, 8 May 2026 (UTC)

 VulpesBot (talk · contribs · logs · block log) (7 May 2026)
Editor has acknowledged AI/LLM use, committed to stop, and will follow WP:NOLLM. Cleanup may still be needed. Test

 Cleanup has been requested.

@Dr vulpes: I like this layout! But we've lost the pencil icon to edit an individual article's status ClaudineChionh (she/her · talk · email · global) 06:37, 9 May 2026 (UTC)

@ClaudineChionh refresh the page and check it out. I think it should work now. I can change the icon to text or something if you think it'll look better. I need to check and make sure it'll work with a screen reader next. Dr vulpes (Talk) 02:32, 10 May 2026 (UTC)

 Johnny Au (talk · contribs · logs · block log) (10 May 2026)
Editor has acknowledged AI/LLM use, committed to stop, and will follow WP:NOLLM. Cleanup may still be needed. 

I would appreciate a second opinion on whether or not these replies to my comment (at the bottom) are LLM-generated. This editor has said they often use an LLM for copyediting, but denies using one to help write these comments. Also, their last reply to me suggested the answer to Q4 was in fact written by an LLM, and it seems they realized this and changed it (see this diff). They also changed their answer to Q4 with the edit summary "rewrote #4 to be human-written." I would like to be proven wrong, but the language used (particularly in their second reply) is just too suspicious for me to let this go. I₂Overcome ^talk 22:11, 10 May 2026 (UTC)

Btw, if you want to link to someone without pinging them, you can use {{noping}} or {{noping2}}, or Special:Contributions/ExampleUser for their contribs Kowal2701 (talk, contribs) 23:19, 10 May 2026 (UTC)

I have decided to withdraw from the May 2026 Admin Election as I felt I wasn't ready yet. Thank you so much for identifying a major weakness in the candidacy. I must admit that the vast majority of the write-up was AI-generated and I must apologize. I will return in the next admin election in October 2026 and I will use my own words when writing then. Johnny Au (talk/contributions) 23:33, 10 May 2026 (UTC)

I’m sorry to hear that, I (and no doubt many others) was planning on voting yes. Well done for your honesty and best of luck in October's elections Kowal2701 (talk, contribs) 23:38, 10 May 2026 (UTC)

Thank you for your honesty. If you were this forthcoming right away, you might still have been successful. Good luck in October. I₂Overcome ^talk 23:42, 10 May 2026 (UTC)

I agree, best of luck next time.

Try to avoid falling down the hole that I see many editors fall into, in which they can no longer function without AI help. Given that AI companies are paying the Wikimedia Foundation for the privilege of training their models on Wikipedia's data, it makes no sense at all for AIs to be trained on their own output. Therefore we must be vigilant in stamping it out when we see it, whether it's in article prose or in editor communications.

Remember, when you are an administrator, you will encounter situations where you must decide whether to block or protect based on the AI usage being reported, and it isn't an easy job. ~Anachronist (who / me) (talk) 23:46, 10 May 2026 (UTC)

To make amends, I have added an extra piece of advice in Wikipedia:Advice for admin elections candidates#Advice for answering questions, this time regarding AI use. It is so others would not fall into the same trap as myself. I will return in October, but this time, it will be written using my words. Johnny Au (talk/contributions) 00:51, 11 May 2026 (UTC)

 Bronchalinox (talk · contribs · logs · block log) (12 May 2026)
Editor has been warned about the use of AI/LLM content and WP:NOLLM. 

CleanEdinburgh of the Seven Seas  reverted

AfDArtificial intelligence in Europe

From this edit (permalink made by Bronchalinox either a paraphrase made by an ai or complete ai generated edit. Clue for being AI generated is the random bolded text, source 2 to 5 leads to a 404 error despite being accessed literally a day ago. Source 11 leads to a different article and different website that stated in the source. Source 13 is a 404 error. Warm Regards, Miminity (Talk?) (me contribs) 11:05, 12 May 2026 (UTC)

I forgot to add that this user’s draft was repeatedly rejected or declined for being AI generated Warm Regards, Miminity (Talk?) (me contribs) 11:09, 12 May 2026 (UTC) I also reverted the edit now. 11:10, 12 May 2026 (UTC)

Besides the clear signs you've identified (not to mention the addition of a new Geography section and a broken ref), I have trouble believing the polished prose was written by the person who wrote that user page. I'm now looking into Artificial intelligence in Europe which was very recently accepted by Nixleovel. ClaudineChionh (she/her · talk · email · global) 11:48, 12 May 2026 (UTC)

Base on this permalink. Yeah the article is at least partially created by AI. Warm Regards, Miminity (Talk?) (me contribs) 12:08, 12 May 2026 (UTC)

I initially went back and forth on that one, the revision history threw me off a bit, but looking at it again, I completely agree that it has the signs of AI-generated content. I’m sorry for the oversight on my part. Regarding broken references, I remember getting a bunch of 500 internal server errors so I assumed the websites may have been down. Thank you for the second set of eyes and for catching that. Nixleovel (He/They) (Talk • Contribs) 19:28, 12 May 2026 (UTC)

 DanieleJava (talk · contribs · logs · block log) (11 May 2026)
This matter has been resolved; no further action is required. 

Requesting a second opinion on whether DanieleJava (talk · contribs · count) is posting LLM-generated comments without disclosure in the Michael Jackson contentious topic (WP:GS/MJ). I initially warned DanieleJava for LLM misuse based on the structure and formatting of their initial RSN comment (context), but retracted the warning after considering the possibility that DanieleJava simply had a writing style that resembled LLM outputs. However, DanieleJava later made an edit that inserted a ChatGPT-generated citation, so I had to reissue an LLM warning.

In the discussion Talk:1993 Michael Jackson sexual abuse allegations § Multiple problems with this article, DanieleJava's initial 8,420-character comment was clearly copied and pasted into Wikipedia, and is formatted with multiple signs of AI writing, including WP:AIBOLD, WP:AILIST, and the peculiar use of Roman numerals ("There are two articles from Charles Thomson (xxxviii)(lxxvi)", "Aphrodite Jones is cited as source (LXVIII)"). This does look LLM-generated to me, and I am seeking additional input from this noticeboard to determine an appropriate response. — Newslinger talk 11:12, 11 May 2026 (UTC); struck portion about Roman numerals 12:01, 11 May 2026 (UTC)

I literally copied the numerals from the bottom of the original article and I always write like using bullet points and formatting, as you can see from a 2023 post in the talk page. The fact that I copied a citation from an article (and I checked its content, as you can see from our debates on the talk page), doesn't mean I am using an LLM. And no, this comment you refer to, is not written by an LLM or with the help of an LLM.

This is an article I wrote (in Italian) in 2010. No LLMS, tons of citations and same formatting. DanieleJava (talk) 11:56, 11 May 2026 (UTC)

Fair enough about the Roman numerals. The rest of the comment still stands. In particular, where did the ChatGPT-generated citation come from if you are not using an LLM? — Newslinger talk 12:02, 11 May 2026 (UTC)

What kind of evidence do you need, more than a 2010 article where I use bold, bullet points, arrows, citations and structured paragraphs? Maybe a second article from 2012? DanieleJava (talk) 12:05, 11 May 2026 (UTC)

An explanation for the ChatGPT-generated citation would be a good start. — Newslinger talk 12:06, 11 May 2026 (UTC)

I swear I have no idea where the GPT citation came from, because I've been opening hundreds of tabs to find the requested evidence for each claim in that article, since any source provided was deemed biased, banned or whatever. I currently have 135 tabs open in Chrome. DanieleJava (talk) 12:08, 11 May 2026 (UTC)

That's not an explanation. ChatGPT utm tags aren't added unless ChatGPT was used. It would help your case significantly if you start with being honest. Katzrockso (talk) 13:53, 11 May 2026 (UTC)

What I'm saying is that I might have copied the citation from God only knows what page, whose author had probably used ChatGPT. My writing style matches and that's literally the only citation with this problem. With all the edits I've made these days, there should be dozens of links. What happened to the WP:AGF all of a sudden? DanieleJava (talk) 18:35, 11 May 2026 (UTC)

I’m actually inclined to WP:AGF here (at least with respect to the LLM allegations; no comment on anything else). This user's comments do not read as LLM-generated to me. The use of bolding and lists alone does not automatically indicate that something is AI; I have been guilty of jumping to this conclusion myself before only to discover that the content in question was written like 15 years ago. As for the ChatGPT citation: unfortunately, lots of things online are being written with LLMs. It is possible that the URL was copied from some other tertiary source without noticing the identifier at the end. I₂Overcome ^talk 18:59, 11 May 2026 (UTC)

Yeah this is weird, on the one hand nothing leaps out in the prose, and this 2023 writing doesn't sound like AI and certainly not AI in 2023. The talk page comments read nothing like AI whatsoever either, sorry. AI does bolding in a very specific way and this isn't it.

As for the AI param: It is finicky to search for this especially given that Google sucks now, but I haven't been able to find any tertiary source containing that URL with that UTM parameter, though -- Google, Bing, DuckDuckGo. Indexing also kind of sucks for this. (I also haven't found any search terms that would lead to Google indexing the URL with the param, which happens once in a blue moon). Gnomingstuff (talk) 07:06, 12 May 2026 (UTC)

IMO neither Special:Diff/1353029953 nor Special:Diff/1353535613 appear to be LLM generated based on stylistic evidence. Although they have some superficially LLM-like structures (bulleted lists, bolding) they don't use those features the way LLMs do, and both edits also contain numerous constructions that IMO are very unlikely to be produced by an LLM. I did not see the chatgpt-tagged citation, could someone share precisely what citation that is referring to? Perhaps Danieljava could then dig through their tabs to find where they copied that citation from, which would be compelling evidence in their favor. -- LWG ^{talk (VOPOV)} 19:08, 11 May 2026 (UTC)

This diff, second to last citation. I₂Overcome ^talk 19:35, 11 May 2026 (UTC)

It seems unlikely to me that copy/pasted chatbot output would include the gpt tag in one LA Times link but not another. The only way I could see chatbot output resulting in this diff would be if DanieleJava edited it to remove AI signs and happened to miss one. -- LWG ^{talk (VOPOV)} 19:53, 11 May 2026 (UTC)

I'd also like to point out that I often had to start writing my posts all over again, because after a while browsing other pages, Wikipedia just decides to delete anything I was writing (or maybe it's Chrome, to save RAM, I don't know). DanieleJava (talk) 20:24, 11 May 2026 (UTC)

I'm leaning not AI for these particular edits but it's fairly common actually for only a few URLs to have the param especially if articles are being worked on or cited piecemeal Gnomingstuff (talk) 07:07, 12 May 2026 (UTC)

It could also be AI assistance to find sources. I have had some success with that Katzrockso (talk) 23:33, 12 May 2026 (UTC)

Amongst the open tabs there is nothing about Sawyer. I don't know if it was a link in some forums or Reddit or anything else. There was (actually, there is) a tense discussion in a couple of talk pages and this led to a frenetic search for anything that could fit in the "requirements" made by the other users and that meant to open several articles, forum pages, reddit posts, blogs and so on and check if the information reported was reliable/sourced. I see that in my browser history there are two pages for that article, both without the gpt thing, opened in different timeframes (plus my attempt to google the URL). Maybe the original page where the link was found had another name. Also, if I put the url with the GPT ending in the search bar, the GPT appendage disappears (?). No idea if this helps.

I do use ChatGPT, but definitely not to write nor research nor anything of the sort. I use it mainly to fix my computer. DanieleJava (talk) 20:21, 11 May 2026 (UTC)

@User:Newslinger does this satisfy your concerns about possible LLM use here? -- LWG ^{talk (VOPOV)} 20:30, 11 May 2026 (UTC)

Writing samples from before the LLM era are one of the strongest forms of evidence for one's natural writing style, so while the addition of the ChatGPT-generated citation is strange, DanieleJava has done the best they can to provide counterevidence here. The WP:GS/MJ contentious topic, for some reason, is a topic area in which LLM misuse is unusually prevalent (see last year's DRN case for example, which includes what appear to be LLM-generated comments from four different editors, only two of whom were identified as LLM users at the time), but DanieleJava has done a decent enough of a job to show that their comments are not like those of the others. — Newslinger talk 19:34, 12 May 2026 (UTC)

Thanks for providing this, this is already like 1000% more transparency than anybody does and I don't think any of this really sounded like AI in the first place. Maybe whatever site you visited before the Los Angeles Times article had that link? Gnomingstuff (talk) 07:11, 12 May 2026 (UTC)

That's my theory as well, but it's tragic to find which one and check every single link or comment. I've provided proof that I have

1. never used any GPT citation in over 3 years of activity on Wikipedia
2. always written like that, with bold fonts, bullet points, lots of citations, long paragraphs and so on

I do understand this is important for you guys, but after all I'm going through for some simple edits on a MJ page, I don't really have the mental energy to debate this too. I leave it up to you to decide and I rest my case. DanieleJava (talk) 11:25, 12 May 2026 (UTC)

I understand -- I think everything does check out here. Gnomingstuff (talk) 12:40, 12 May 2026 (UTC)

Suppose an editor asked (say) ChatGPT to locate sources on (say) P=NP. That’s fine: it's a free country. ChatGPT comes back with some interesting papers. That’s fine: we’ve been doing this with Google for eons. Later, the editor adds a comment to Talk:Computational Complexity quoting one or these papers and including the URL with the ChatGPT campaign utm. I cannot see why this is transgresses policy. MarkBernstein (talk) 15:37, 12 May 2026 (UTC)

It’s isn’t against PAGs to get links from LLMs, but there’s an overlap between people who do that and those that use LLMs to generate content, so it invites greater scrutiny Kowal2701 (talk, contribs) 16:04, 12 May 2026 (UTC)

There is nothing wrong with using an LLM to conduct research. However, when an editor who inserts an LLM-generated URL says they have not used LLMs for Wikipedia at all, their claim needs to be evaluated by examining their other edits. Even if these other edits seem acceptable, it is still worth making an effort to trace the origin of the LLM link, which may lead to LLM misuse by another editor on another Wikipedia page. — Newslinger talk 19:45, 12 May 2026 (UTC)

I mean, I tend to correct bachelor thesis in Italian (Austrian students who learn Italian at university) and some have citations made by some kind of AI. Since my job includes proofreading them, I do understand where the issue comes from. Most of them are totally made up (links don't bring anywhere) and those that aren't, do not actually contain the piece of information that the sentence should back up. I don't even know why these people attend university, at this point. DanieleJava (talk) 22:03, 12 May 2026 (UTC)

N Cleanup is unnecessary.

 BobbyJSmith1972 (talk · contribs · logs · block log) (12 May 2026)
This issue is currently under review. 

Despite denying LLM usage, this editor appears to be using AI on a draft as well as many articles over the past year. BobbyJSmith1972 claims to not be using LLMs even though their draft Draft:Matt Shaffer has false claims (I noted one of them here, which they then removed with edit summary "Removed ESPN article due to reviewer having issue") and at least one reference that leads to an error page (e.g. ref #35 in this version). Note that the user talk page responses also appear AI-generated.

Looking through their other contributions, they have inserted AI text into many articles through newcomer tasks. e.g. this diff has vague phrasing like "Vento Motorcycles U.S.A. was described as having its headquarters in San Diego" and "The brand’s history has been linked to U.S. and Mexico operations". This has been going on for many months ("Recent research on open-source geospatial content-management solutions highlights their use for heterogeneous data integration, interactive mapping and cross-platform publishing in web and mobile environments") Helpful Raccoon (talk) 04:41, 13 May 2026 (UTC)

Some of their earlier edits were reverted by @Gnomingstuff. Helpful Raccoon (talk) 05:13, 13 May 2026 (UTC)

Again, I would request a formal review of this situation with @Helpful Raccoon under WP:HARASS, WP:AGF and WP:BITE. BobbyJSmith1972 (talk) 05:28, 13 May 2026 (UTC)

I am a real person and find the assumptions insulting. These repeated accusations by @Helpful Raccoon (despite my repeated clarifications) are reaching a level of harassment (WP:HARASS) and bad-faith (WP:AGF) behavior. I do understand the need for quality control from other volunteers such myself, but constant challenging a contributor's identity simply because they use formal writing and proper structure is not how this should be handled. If Wikipedia continues to 'bite' and punish contributors for professional writing styles, the platform will become significantly less useful. I am happy to address the comments raised by this contributor. I simply removed the ESPN source stated to de-escalate the situation as I felt attacked. Also the edit to Vento was a mistake as I used a word document I use to draft my edits. It was clearly a mistake of copy and paste. I will not make that mistake again. I truly thought the newcomer tasks are a good way to contribute and build rapport within this community. I appears that @Helpful Raccoon wants to punish those who want to improve Wikipedia for others. BobbyJSmith1972 (talk) 05:16, 13 May 2026 (UTC)

I apologize that I didn't link the policies properly in my previous message out of frustration. I would like a formal review of this situation under WP:HARASS, WP:AGF, and WP:BITE. BobbyJSmith1972 (talk) 05:25, 13 May 2026 (UTC)

I'm not sure how an edit's vague phrasing could possibly be explained by "a mistake of copy and paste". Helpful Raccoon (talk) 05:43, 13 May 2026 (UTC)

No one is challenging your identity. They're asking questions about the tools you are using. Gnomingstuff (talk) 12:43, 13 May 2026 (UTC)

Reported to ANI. Helpful Raccoon (talk) 07:23, 13 May 2026 (UTC)

Btw @Helpful Raccoon, you can use {{@AINBA}} to avoid having to go to ANI Kowal2701 (talk, contribs) 11:09, 13 May 2026 (UTC)

I don't mean to get in the way of this whole thing, but I'm giving my own insight at the draft provided.

Citations 1, 5 and 8 doesn't reference the subject at all. Citation 12 doesn't really explicitly name the sponsor and same with citation 13. The couple excessive citations should be noted or actually cite their respective claims instead of putting it at the end of the paragraphs.

I've only glanced a little but I'll be sure to give a hand at reviewing the draft if anyone wants a third party to help (even though I don't actually hold authority at what's accepted or not).

PeepeeDino (talk) 11:03, 13 May 2026 (UTC)

 Habeshakira (talk · contribs · logs · block log) (12 May 2026)
Editor has been warned about the use of AI/LLM content and WP:NOLLM. 

CleanBattle of Adwa  Already reverted

Needs reviewEritrean–Ethiopian War

Needs reviewMedri Bahri

Needs reviewNaýib Al-Bahr

Needs reviewDäbbäb Arýaya

Needs reviewOromo expansion

Needs reviewYohannes IV

Needs reviewRas Alula

Needs reviewProvinces of Eritrea

Needs reviewZämäča

Needs reviewEthiopian National Literacy Campaign  I2Overcome  I tagged and draftified it, but the author untagged it, moved it back, and then a NPR immediately reviewed it, so...

Needs reviewEducation in Ethiopia

CleanKingdom of Aksum  I2Overcome  Reverted edit that removed numerous refs

Moxy🍁 02:40, 13 May 2026 (UTC)

I'm going to need another opinion on this. It's possible this isn't actually LLM use, but there are some signs such as redlinked cats, mentioning the source in the prose, and focus on significance. I₂Overcome ^talk 05:32, 13 May 2026 (UTC)

I’ve got access to the source they’re using so will have a look content-wise. Their userpage is reminiscent of LLM-generated ones, but their edits seem okay Kowal2701 (talk, contribs) 09:33, 13 May 2026 (UTC)

Nvm, the copy I have is botched and it would take me hours to download each volume. FWIW I don't think this editor is using AI, they strike me as very competent on the topics Kowal2701 (talk, contribs) 11:06, 13 May 2026 (UTC)

Yup its huge https://journals.sub.uni-hamburg.de/aethiopica/article/view/294 Moxy🍁 13:25, 13 May 2026 (UTC)

I use athiopica for all my edits . If you would like for any reference I’ve made I can send you a picture or quote the verse of the reference. I believe that ai likes to hallucinate a lot and provide false informations . All of my references are 100% correct . Habeshakira (talk) 15:20, 13 May 2026 (UTC)

They've said they used AI for their userpage but not for editing in mainspace . I believe them, they come across as very genuine, this can probably be marked as "cleanup unnecessary" unless anyone thinks otherwise Kowal2701 (talk, contribs) 15:32, 13 May 2026 (UTC)

 KurdishPAL (talk · contribs · logs · block log) (13 May 2026)
This issue is currently under review. 

Hi all, KurdishPAL is using AI to generate articles as evidenced by oaicite tags in their citations.

I've just nominated Battle of Sidi Khalef for G15 since it has the AI tag and precisely one source - page 338 of a book that doesn't have 338 pages. I've even looked on Google Books in case the AI got it from there, but the 338th scanned page is blank. The article is also using Wikipedia as a source, something else that happens with AI.

I can see a notice for a previous draft that was declined for AI-use , but this editor is now unfortunately bypassing AFC and creating AI-generated articles directly into mainspace.

Can someone please take a look at their previous edits and see whether anything else has been missed? Blue-Sonnet 16:02, 13 May 2026 (UTC)

Hello @Blue-Sonnet

I would like to clarify that I translated the article from the French Wikipedia page. I did not intend to create fabricated content or bypass review. However, I accept that the page I’ve translated has sourcing problems.

I will move the page and I will remove the problematic citation and trim any unsupported claims, and improve the article with reliable sources.

Sorry for the issue. I will be more careful with citations and article creation in the future. KurdishPAL (talk) 16:12, 13 May 2026 (UTC)

Hi @KurdishPAL, there's only one paragraph of text in the French article, where did the rest of the article come from? It would be great if you can use WP:AFC going forwards. Blue-Sonnet 16:15, 13 May 2026 (UTC)

Hi, I translated the French article as the basis, but I also followed the same layout as the previous battle articles I had created, so I expanded it into sections like Background, Battle and Aftermath. I now realise that I shouldn’t have added material beyond what was clearly supported by the sources. That was my mistake, and I’m sorry, But as I said, the French page contains the source you said was an AI hallucination. KurdishPAL (talk) 16:20, 13 May 2026 (UTC)

Ok, can you please agree to stop using AI, LLM's or chatbots to generate articles or sources?

I can see that the additional text you added was AI-generated & that's against the Wikipedia:NEWLLM guidelines. You should also manually check every source, even if it's from another article - Wikipedia:Translation and Wikipedia:LLM-assisted translation give you instructions to follow.

I highly recommend using AFC until you've had multiple drafts accepted (maybe ten or more), that way you can be sure you're making good articles.

Please be very careful from now on, if you continue to use AI/LLM/chatbots then you might face administrator sanctions because of the many problems it causes. It's honestly best not to use AI at all here. Blue-Sonnet 16:45, 13 May 2026 (UTC)

I understand, and I appreciate the clear guidance. I’ll stop using ai KurdishPAL (talk) 16:48, 13 May 2026 (UTC)

 AkhiXyz (talk · contribs · logs · block log) (13 May 2026)
Referred to Administrators' Noticeboard/Incidents. First warning, additional warnings, ANI.

CleanOrthohantavirus  Ca

CleanClarence Darrow  RandFreeman

CleanMental health of Jesus

CleanMissing scientists conspiracy theory  Ca

CleanSaturn  Em-as-in-emily

CleanSolo Leveling  Ca

CleanTobacco and Vapes Act 2026  Ca

CleanYantra  Ca

CleanThe Loud House  Ca

User:AkhiXyz (~70 edits) shows clear signs of AI usage (exceedingly verbose edit summaries), resulting in source-text integrity issues , and close-paraphrasing . I've warned them for now, but their contributions need a closer look. Ca ^{talk to me!} 15:14, 6 May 2026 (UTC)

Reported at Wikipedia:Administrators' noticeboard/Incidents#AkhiXyz -- possible use of AI. --Guy Macon (talk) 12:32, 14 May 2026 (UTC)

 AnilDiggiwal (talk · contribs · logs · block log) (14 May 2026)
Editor has been notified and asked to discuss AI/LLM editing concerns. 

AnilDiggiwal (talk · contribs · deleted contribs · logs · filter log · block user · block log)

Needs reviewBhagwatilal Rajpurohit

Needs reviewFred Negrit

Needs reviewGafruddin Mewati Jogi

Needs reviewHarchandan Singh Bhatti

Needs reviewLalsot

Needs reviewMuni Narayana Prasad

Needs reviewNavjivan Rastogi

Needs reviewNeelkanth Mahadev Temple, Dausa

Needs reviewOnkar Singh Pahwa

Needs reviewSridhar Makam Krishnamurthy

Needs reviewSurendra Kishore

Needs reviewSwami Brahmdev Ji Maharaj

Needs reviewTaga Ram Bheel

Needs reviewVeezhinathan Kamakoti

Needs reviewVikrant University

Needs reviewZahir Ishaq Kazi

Has created since 2025 several articles that appear to be AI-generated, most recently yesterday. The articles are fairly clear WP:AISIGNS-wise in terms of prose, and initial diffs are even more pronounced in this, such as this Markdown formatted 2025 draft. The AI traits are also consistent with the dates the articles appeared -- AIATTR shows up more pronouncedly in the later stuff, etc. This obviously does not apply to the edits from 2022. Gnomingstuff (talk) 03:25, 14 May 2026 (UTC)

Added an initial semi-random list, there may be more but I'm trying to maintain some sanity. ClaudineChionh (she/her · talk · email · global) 12:53, 14 May 2026 (UTC)

 Cleanup has been requested.

MaynardClark (talk · contribs · deleted contribs · logs · filter log · block user · block log)

I opened this last year with the comment: Fairly prolific with edits, having the usual WP:AISIGNS all over the place going back to 2023; but what elevates this to an actual noticeboard post is the talk page comment here, User_talk:MaynardClark#ChatGPT, specifically the last paragraph which suggests that review isn't really being done.

Well, they're still making articles, they still appear to be AI (and, like the talk page comment above mentions, they apparently disclosed by email that they were using AI, which obviously I can't confirm but WP:DUCK applies), and they still don't seem to be acknowledging anything.

Also, sorry for the multiple reports, I am slogging through the 6,000 talk page sections I have to create because people can't just believe you about AI without your signing forms in triplicate, and that is eating up all my time Gnomingstuff (talk) 07:09, 15 May 2026 (UTC)

Up to you, but imo there ought to be more reports, sanctioning editors who fly past warnings saves us from doing more cleanup, and even if the cleanup doesn't get completed, it’s good to have it documented and an archive to work through. Just tagging and cleaning up easy-to-detect articles without focusing on users would let a lot pass under the radar Kowal2701 (talk, contribs) 10:44, 15 May 2026 (UTC)

Moved from Wikipedia:Village pump (miscellaneous)/Archive 88#Possible LLM use found and then Wikipedia:Village pump (policy)#Possible LLM use found

 – George Ho (talk) 17:17, 27 May 2026 (UTC)

On the article Wonderoos, I saw that the writing style is very reminiscent of an LLM, especially in the Premise section. Also there are mistakes, like calling the in-universe pet species (Poofy Schmoop) "Schmoopsy" I've added Template:AI-generated to the article, but is it okay if you could review it? RotatingPirateShip (talk) 15:45, 27 May 2026 (UTC)

• WP:LLM is a guideline, so moving this thread by RotatingPirateShip to this Village Pump page then... —George Ho (talk) 16:51, 27 May 2026 (UTC)

  @George Ho May I ask why you moved this thread here? The banner at the top says For questions about how to apply existing policies or guidelines, refer to one of the many Wikipedia:Noticeboards. That seems to apply here – it seems WP:AINB might have been a better venue. Toadspike [Talk] 17:07, 27 May 2026 (UTC)

  ...Oh... Didn't know that. Will do so soon. —George Ho (talk) 17:13, 27 May 2026 (UTC)

  Moved. RotatingPirateShip (talk) 18:35, 27 May 2026 (UTC)

  Disregard comment above. But, this is AINB. RotatingPirateShip (talk) 18:36, 27 May 2026 (UTC)

It also looks LLM to me. The plot summaries are entirely unsourced--that is a LOT of unsourced text. The statement that What-To-Doodles became Wonderoos is unsourced (neither of the sources bracketing it cover this). The citations in "Premise" are extremely weirdly placed in the middle of a sentence: if that's interpreted normally, the rest of Premise is unsourced. "Characters" is a ton of material almost completely unsourced. Where is all this stuff coming from? Probably LLM, either used here or used in a source and copy-pasted from that source without citation. I am not sure this can be cleaned up without reducing it to a stub, frankly. M kuhner (talk) 23:12, 27 May 2026 (UTC)

Yes. Even though What-To-Doodles's theme appears to be similar to Wonderoos, there isn't really any official confirmation from Netflix that What-To-Doodles was a precursor to Wonderoos. This article should be COMPLETELY rewritten. RotatingPirateShip (talk) 15:04, 28 May 2026 (UTC)

Disregard post above. I personally think the page should be deleted. Nominating for AfD. RotatingPirateShip (talk) 15:29, 28 May 2026 (UTC)

AfD'd. See Wikipedia:Articles for deletion/Wonderoos RotatingPirateShip (talk) 16:56, 28 May 2026 (UTC)

Just a note here -- plot summaries don't need sources as they are assumed to be sourced to the story/film/etc itself. Obviously LLM-generated plots are not allowed, but it's not because of lack of sourcing. Gnomingstuff (talk) 20:34, 29 May 2026 (UTC)

Thanks, I did not know that. M kuhner (talk) 21:25, 29 May 2026 (UTC)

 Daniel G Rego (talk · contribs · logs · block log) (31 May 2026)
Editor has been blocked site-wide for AI/LLM policy violations. ANI

Daniel G Rego (talk · contribs · deleted contribs · logs · filter log · block user · block log)

See - Here is the full copy-paste-ready version with the explanatory notes applied throughout the article. I also cleaned the duplicate FUNAGSecretaries reference definition and fixed the “Thi is” typo in the major officers section. Cleanup required. Fermiboson (talk) 11:45, 28 May 2026 (UTC)

Hi Fermi,

Thanks for the flagging, and thanks for creating the bot to check for suspicious edit summaries.

I edit the article myself and take full responsibility for the content, sourcing, and final wording.

Before publishing major changes, I check for syntax problems, reference formatting, disambiguation issues, date consistency, MOS-style cleanup, and similar issues. However, to your concern, there's no machine-generated content in the article, nor any unsourced, knowingly untrue, promotional, non-neutral, etc or otherwise content.

I appreciate you keeping an eye out, though. PLEASE keep up the good work!

Cheers,

Dan Daniel G. Rego (talk) 12:05, 28 May 2026 (UTC)

Please stop lying to our faces given that you managed to paste communication intended for the user from the LLM into the article. Fermiboson (talk) 12:07, 28 May 2026 (UTC)

Hey Firmi - please keep civility. We are too old and busy to be in never-ending ping-ponging on internet boards! =) Again, I take full responsibility for the article’s content, sourcing, and final wording. Like all Wikipedia articles, it ultimately rests on the sources, and the article is tied to them. But again, thanks for your time, and if there's anything else, specifically that requires action, please reach out here or via discord. Otherwise, I gotta go take my daughter now to class. Thanks! Daniel G. Rego (talk) 12:21, 28 May 2026 (UTC)

While my fellow editor's tone was out of line, the content concerns are valid. The fact you copied and pasted communication intended for the user into the article, even in error, means we can't trust that the edit matches what the sources say. When in your editing process do you make the checks for content and syntax? SenshiSun (talk) 14:06, 28 May 2026 (UTC)

While my fellow editor's tone was out of line, what? I suggest reading the civility policy and noting that while nothing Fermiboson said is in there, WP:LYING is, and Daniel G. Rego has lied. fifteen thousand two hundred twenty four (talk) 20:12, 28 May 2026 (UTC)

Agreed. There's nothing wrong with what Fermiboson said, or how they said it. --gurkubondinn 06:59, 30 May 2026 (UTC)

Noting that I left similar concerns on this user’s talk page earlier this month, which they also denied. JTtheOG (talk) 19:36, 28 May 2026 (UTC)

I take full responsibility for the article’s content, sourcing, and final wording – Inadequate, you may not generate article text at all, period. It doesn't matter what the sources are or how much responsibility you take, the use of LLMs to generate or rewrite article content is prohibited per WP:NOLLM.

You have made aware of this twice before. Stop generating text and stop lying about generating text. fifteen thousand two hundred twenty four (talk) 20:18, 28 May 2026 (UTC)

Isn't lying twice about it grounds for sanctions? JTtheOG (talk) 20:39, 28 May 2026 (UTC)

It is a case of repeatedly breaking the rules (sanctionable), knowing he has repeatedly broken the rules (exacerbating), AND repeatedly being less than honest with his fellow editors about it (incivility).

I am quite taken aback that Daniel G. Rego has decided to continue denying using an LLM for content despite there being 100% solid proof that he did (leaving in the message from the LLM in his cut and paste). In the light of this, any defence of taking responsibility for one's edits and of supposedly assiduous checking fall apart because we can no longer trust such claims.

The no LLM policy is there because the community has decided their use for content causes far more problems than it is worth (and this noticeboard keeps turning up evidence why). If an editor disagrees with that policy, they need to take it up in the right forum, rather than engage in disruptive behaviour. OsFish (talk) 04:31, 29 May 2026 (UTC)

@Daniel G Rego: can you read the WP:USED-AI? essay, and then let us know if you might want to revise your answer? --gurkubondinn 00:03, 30 May 2026 (UTC)

Hi Gurku, I got your notification. Didn't realize there was further conversation on this thread since my last post.

I want to clarify my process here:

My understanding of WP:LLM/WP:NOLLM is:

"Editors are permitted to use LLMs to suggest basic copyedits to their own writing, and to incorporate some of them after human review, provided the LLM does not introduce content of its own. Caution is required, because LLMs can go beyond what is asked of them and can change the meaning of the text such that it is not supported by the sources cited. Examples of basic copyedits include spelling, punctuation, and capitalization."

Therefore, after the work is performed, using assistance such as:

I need a final pass, line by line of my revision to this article against the latest English Wikipedia guidance, including WP:LAYOUT, WP:CITE, MOS:REFPUNCT, MOS:DATEUNIFY, WP:DATERANGE/MOS:DATERANGE, WP:NBSP/MOS:NBSP, WP:NDASH/MOS:NUM, citation style consistency, citation order, named-reference syntax, correct placement of explanatory notes, duplicated references, missing or undefined references, unused references, malformed URLs, template syntax, table syntax, image syntax, list formatting, section order, categories, DEFAULTSORT, hatnotes, portals, navboxes, and any other syntax, typo, formatting, or layout issues you find. Do not omit sections, images, references, lists, tables, or categories unless there is a clear English Wikipedia policy reason to do so, and explain any such omission.

In order to catch technical and style issues and similar cleanup issues. Therefore, the article’s content, sourcing, organization, and final wording are my full responsibility and I believe the revision improved the article from the state in which I found it.

But I understand why patrollers are being careful here. That's actually why I love reading Wikipedia, and (when I can, volunteer my time and money), because the quality here tends to be quite good.

And I get, I've been here for a long time, and I've seen my fair share of vandalism, low-quality editing, and all sorts of disruptions your team has to deal with it.

So, again I appreciate the concerns, but that being said, I rest my case.

Have a nice weekend!

- Cheers,

Dan Daniel G. Rego (talk) 00:59, 30 May 2026 (UTC)

@Daniel G Rego, can you explain why the prompt you provided above meets the definition in Examples of basic copyedits include spelling, punctuation, and capitalization? —In solidarity with Wiki Workers United · ClaudineChionh (she/her · talk · email) 04:01, 30 May 2026 (UTC)

@Daniel G Rego: so you are conceding that you used an LLM? --gurkubondinn 07:03, 30 May 2026 (UTC)

Note: Just found a draft in my notes that I was gonna post here:

This user has made a series of edits which I found concerning, publishing a rapidfire list of huge articles (or article re-writes) within 20-30 minutes of eachother, showing signs of AI use (such as WP:OVERATTRIBUTION, WP:AILEGACY, and WP:AISUMMARY). This concern was brought to their talk page on May 13 by User:Boynamedsue, who got no response. Following more creations, I asked more directly the next day and they claimed to only use "a bot" to create citations.

His "translations" of articles sometimes look like they've been ran through a chatbot. For example, the pt.wiki version of pt:Afrânio de Melo Franco mentions he was nominated for a Nobel Prize, while the English translation explains that his Nobel Peace Prize nominations in the 1930s reflected contemporary recognition of his involvement in diplomacy. JTtheOG (talk) 22:44, 28 May 2026 (UTC)

I'm also seeing some signs in Parque Municipal Américo Renné Giannetti, mainly The city has also listed Tai Chi Chuan classes at Praça do Trenzinho among the park's regular free activities. ... Environmental education also appears in the park's public calendar. + red see also links in the current revision. Also in that article's creation edit summary: Created article on Parque Municipal Américo Renné Giannetti, translating and substantially adapting ptwiki article with expanded history, features, cultural spaces, public art, events, gallery, references, further reading, Commons link, authority control, portal bar and English Wikipedia categories. Added official, heritage, cultural and institutional sources. Please refer to Talk Page for further details.

Sarsenet•he/they•(talk) 12:00, 29 May 2026 (UTC)

Reported at Wikipedia:Administrators' noticeboard/Incidents#AI use by User:Daniel G Rego. We should spend as much time humoring these plagiarists as they do writing in their own words. Thebiguglyalien (talk) 01:30, 31 May 2026 (UTC)

An LLM-suggested edit summary was also accidentally pasted into article space (albeit inside HTML comments) at the bottom of this diff, which seems to have involved LLM translation from Portuguese but also created additional material, such as a "Selected Works" section whose citations are all to Wikimedia Commons entries. WP:LLMTRANSLATE is a bit laconic, and charitably one might think it gives permission to prompt an LLM with something loose like "write an English Wikipedia article using the text of the Portuguese article on the same subject", rather than a narrow instruction like "translate the following paragraphs into English from Portuguese". To my mind it would be better to clarify that only the latter is allowed. Einsof (talk) 03:03, 31 May 2026 (UTC)

Cleanup table

LLM PRODSerra do Curral

LLM PRODParque das Mangabeiras

LLM PRODPaul Villon

LLM PRODEdifício Acaiaca

LLM PRODPalácio das Artes

LLM PRODBrazilian Academy of Music

LLM PRODMuseu da Música de Mariana

LLM PRODDanaïde (Brâncuși)

LLM PRODAarão Reis

LLM PRODOrquestra Ribeiro Bastos

LLM PRODOrquestra Lira Sanjoanense

DeletedPedro Miguel de Almeida Portugal e Vasconcelos  llmprod

LLM PRODFrancisco Curt Lange

LLM PRODIgnácio Parreira Neves

DeletedCaptaincy of Minas Gerais  llmprod

LLM PRODJosé Tomás Nabuco de Araújo

LLM PRODJosé Tomás Nabuco de Araújo Filho

LLM PRODMaurício Nabuco

LLM PRODCarlos Martins Pereira e Souza

LLM PRODMário de Pimentel Brandão

LLM PRODJosé Carlos de Macedo Soares

LLM PRODAfrânio de Melo Franco

LLM PRODAfonso Arinos de Melo Franco (jurist)

LLM PRODBruno Lechowski

LLM PRODMuseu Casa de Portinari

LLM PRODRua do Ouvidor

LLM PRODCasa Leuzinger

DeletedGeorge Leuzinger  g15

LLM PRODRoyal Military Academy of Rio de Janeiro

LLM PRODJean Léon Pallière

LLM PRODArnaud Pallière

LLM PRODJoão de Deus de Castro Lobo

CleanNational Quarantine Unit  substantial contribution by another editor, ineligible for prod, content removed in Special:Diff/1357976803

M kuhner (talk) 00:36, 1 June 2026 (UTC)

Creations tagged accordingly, edits to existing articles reverted where possible and reasonable. fifteen thousand two hundred twenty four (talk) 09:01, 2 June 2026 (UTC)

 Shree Tin (talk · contribs · logs · block log) (1 June 2026)
Editor has been blocked site-wide for AI/LLM policy violations. 

Y Cleanup has been completed.

CleanBalen Shah cabinet  cleaned by LaundryPizza03

CleanRanodip Singh Kunwar  cleaned by fifteen thousand two hundred twenty four

CleanJung Bahadur Rana  cleaned by Zackmann08

CleanTakam Kingdom  cleaned by fifteen thousand two hundred twenty four

CleanDipendra of Nepal  cleaned by Whyiseverythingalreadyused

CleanEnmebaragesi  cleaned by Whyiseverythingalreadyused

CleanHouse of Representatives (Nepal)

CleanBam Bahadur Kunwar  WP:LLMPRV applied in Special:Diff/1358190860

Clean40-Point Demands

CleanBudhanilkantha School

Shree Tin has made multiple massive and highly complicated additions on the same day (the latter of which they self-reverted then rewrote in a matter of 6 minutes, incredible!), and is continuing today . I spot-checked two:

• Budhanilkantha School – full of hallucinated website refs which are marked as being accessed today, despite not existing today, nor ever according to the internet archive.
• Takam Kingdom – mostly uses print refs which are a pain to verify, but I spot-checked uses of Thapa Politics in Nepal, Ancient Nepal and Medieval Nepal and found pervasive verifiability failures there as well.

This appears to be the result of LLM-generation (after all every self-respecting LLM user simply must add sections concerning cultural legacy and significance ).

Looking to form consensus for presumptive removal. fifteen thousand two hundred twenty four (talk) 20:35, 29 May 2026 (UTC)

Go ahead, everything is garbage. –LaundryPizza03 (dc̄) 22:32, 29 May 2026 (UTC)

Go on then, you have my support. Fortek67 (talk) 22:37, 29 May 2026 (UTC)

 Done –LaundryPizza03 (dc̄) 22:58, 29 May 2026 (UTC)

Support using presumptive removal after having looked at the diffs and user contribs. --gurkubondinn 06:25, 30 May 2026 (UTC)

Should we just block the user at this point to prevent them from causing any more damage? –LaundryPizza03 (dc̄) 00:46, 1 June 2026 (UTC)

Since they were warned they have made no further LLM edits, a block is unnecessary at this time. fifteen thousand two hundred twenty four (talk) 00:54, 1 June 2026 (UTC)

I'm not sure what should be done, but it seems that Shree Tin is the same as Renamed user 7779156af9e40e8b3eb3d500bbd466c6, who was just brought here at WP:AINB § User:Himvat before vanishing. Can you double check?

• EIA
• Timecards:
• Both active on commons and newiki
• Vanish request processed at 12:17, 27 May 2026, Shree Tin created 40 minutes later

NicheSports (talk) 01:41, 1 June 2026 (UTC)

Definitely, the article interactions and timing are beyond coincidence. Restoring edit at Dipendra of Nepal, compare Shree Tin to Renamed user 7779156af9e40e8b3eb3d500bbd466c6 , continuation of edits to the "Cabinet Ministers" section at Balen Shah cabinet, continuing infobox editing at Jung Bahadur Rana. @Admins willing to patrol AINB: Shree Tin has improperly used vanish on a prior account to evade scrutiny in violation of WP:VANISHNOT, requesting unvanishing and for one account to be blocked, preferably Shree Tin so the majority of their contribution history isn't obscured going forward.

Also as this is the same user, I consider the above consensus for presumptive removal to apply to the prior account as well (and their last edit there even has markdown in it ). fifteen thousand two hundred twenty four (talk) 02:04, 1 June 2026 (UTC)

Blocked Shree Tin indefinitely as a sockpuppet of Renamed user 7779156af9e40e8b3eb3d500bbd466c6 (talk · contribs), formerly Himvat (talk · contribs), and also for LLM misuse. Requested unvanishing of Himvat on Meta-Wiki.​ — Newslinger talk 08:58, 1 June 2026 (UTC)

Created clean up tables based on Shree Tin edits. Should I also add Himvat edits here too or wait until unvanishing? - Ivan530 (Talk) 11:36, 1 June 2026 (UTC)

Up to you. I think I've reverted all their larger contributions and prodded any creations. fifteen thousand two hundred twenty four (talk) 14:36, 1 June 2026 (UTC)

All done except Bam Bahadur Kunwar. Could someone with more reversion skills tackle that, please? M kuhner (talk) 04:11, 7 June 2026 (UTC)

Done, marking cleanup as completed. fifteen thousand two hundred twenty four (talk) 04:25, 7 June 2026 (UTC)

 Wikiemjay (talk · contribs · logs · block log) (6 June 2026)
This matter has been resolved; no further action is required. 

• big edit only
• Special:Diff/1351688583 self referencing wikipedia on Forward (association football)
• 3/4 dead link on Special:Diff/1352180300

Lokiretro (talk) 10:00, 6 June 2026 (UTC)

Thank you for your request. It's pretty clear that the edits are purely LLM-generated. I have went ahead with reverting all edits and also sending a message to the user. Fortek67 (talk) 12:07, 6 June 2026 (UTC)

 Beady4093 (talk · contribs · logs · block log) (6 June 2026)
This matter has been resolved; no further action is required. 

I'd greatly appreciate a second pair of eyes on this editor, who appears to be using an API (likely AI) to create articles.

Whilst most of the citations work, they often have the wrong year on or other errors.

Their article text is filled with overused em-dashes, overly complex and AI-typical language and there is often duplication of entire sections of articles - or the entire article, as is the case with Chandrasekhar–Friedman–Schutz instability. They've also created duplicate drafts of existing articles, such as Draft:Gravitational self-force. These two diffs are less than twenty minutes apart:

For more detail about the API suspicion, please see their Talk page and .In solidarity, Blue-Sonnet (I'm listening) 13:51, 6 June 2026 (UTC)

Hi. To me it seems very likely the articles are LLM-generated. Gravitational self-force contains a section title that has an upper-case letter for every word, which is common in LLM-generated articles. After looking at their contributions, the same article was created just two minutes after Draft:Gravitational self-force, both are very large in size. As you've mentioned, r-mode has many en-dashes in the text. Also all refs seem to have very specific names, which is also common in LLM-generated articles. Fortek67 (talk) 14:06, 6 June 2026 (UTC)

I've also just realised the way the references have been cited is very odd. They are all cited under the refs section, were their name is used in the article if they want to be cited. What I just said probably doesn't make much sense so see r-mode in the source editor. All citations are also marked as "open-access", which is a bit odd as they don't need to have that.

As a new pages patroller, I think it's appropriate for me to mark all articles as unreviewed. I'm also tagging them as potential AI, and with any other issues they have. Fortek67 (talk) 14:11, 6 June 2026 (UTC)

That saves me a job, thanks! Here's a list if you haven't got them all yet, do you think any are G15-worthy?

• Draft:Gravitational self-force
• Draft:Continuous gravitational waves
• Chandrasekhar–Friedman–Schutz instability
• r-mode
• Draft:Secular instability

In solidarity, Blue-Sonnet (I'm listening) 14:14, 6 June 2026 (UTC)

Thanks for the list. I'm going to nominate the two articles for speedy deletion. I'll also go ahead with reviewing the submitted drafts. Fortek67 (talk) 14:18, 6 June 2026 (UTC)

Strange, Draft:Continuous gravitational waves says quantum squeezing was used in citation 7 but I'm not finding it in the PDF - are you able to find it? I'm on mobile so it could be a bug on the search function... In solidarity, Blue-Sonnet (I'm listening) 14:27, 6 June 2026 (UTC)

Good find – if you're talking about this PDF, I can't seem to find anything either! Fortek67 (talk) 14:35, 6 June 2026 (UTC)

Rightio, G15 candidate! In solidarity, Blue-Sonnet (I'm listening) 14:44, 6 June 2026 (UTC)

Right, should be done now. Secular instability and r-mode nominated for CSD, and I've also declined Draft:Continuous gravitational waves. All other edits have been reverted. Fortek67 (talk) 14:50, 6 June 2026 (UTC)

Sorry I just realised that Gravitational self-force was actually created by another user and was expanded by Beady4093. I didn't realise their edits were reverted, so that article will be staying. Fortek67 (talk) 14:32, 6 June 2026 (UTC)

Yep, they created a duplicate draft! In solidarity, Blue-Sonnet (I'm listening) 14:34, 6 June 2026 (UTC)

Looks like Draft:Secular instability got accepted today... will nominate this article for CSD. Fortek67 (talk) 14:40, 6 June 2026 (UTC)

I'm surprised any of these got through, every source I've spot-checked fails verification... Thanks for your help! In solidarity, Blue-Sonnet (I'm listening) 14:54, 6 June 2026 (UTC)

Yeah, I don't think these AfCs are reviewed well. Or at least reviewers can't really tell an AI article. The sad thing is that about 70% of AfCs I review now have been generated by an LLM. I do my best in stopping stuff like this. Fortek67 (talk) 14:58, 6 June 2026 (UTC)

It's definitely getting more and more frequent, it's only going to get worse too. In solidarity, Blue-Sonnet (I'm listening) 23:10, 6 June 2026 (UTC)

 AMurmu (talk · contribs · logs · block log) (7 June 2026)
This matter has been resolved; no further action is required. 

Created six articles about five-six minutes apart each: Pride celebrations in Chile, LGBTQ history in Nicaragua, Closer (Dennis Cooper novel), Los fantasmas se cabrearon, De mangos y albaricoques, and Psylocke (2024 comic book).

Closer (Dennis Cooper novel) has some hallucinated quotes: the Kirkus review itself says it's part gallows humor, part pornographic excess, part mad-slasher suspense--and definitely not for the squeamish, while the article quotes the text as part black comedy, part pornographic excess, part demented suspense slasher, and definitely not for the squeamish,. The Publishers Weekly quote listed says painful prose, while the actual review says dolorous prose. The creation of the page Gigi Dutreix is tagged for possible AI-generated citations as well. ScalarFactor (talk) 17:11, 6 June 2026 (UTC)

Hi ScalarFactor, thank you for your request. Looking at the user's xtools page, a total of 17 articles were created in two individual occasions; 8 articles on 1 June, 9 articles on 1 June, plus one more article, Vyachaslau Bortnik, on 2 June. During both occasions, the articles were created during a very short timeframe, just minutes apart.

With that alone it's pretty clear that the articles have been LLM-generated. With your discovery of the AI hallucinations and 404 citations, that fully confirms all articles are LLM-generated. For instance, Psylocke (2024 comic book) ref 1 leads to a 404 page. The access date of all citations is incorrect. The first diff has a red-linked article.

I'm going to go ahead with the mass nomination for the deletion of all articles, per WP:LLMPRV. Fortek67 (talk) 00:01, 7 June 2026 (UTC)

Another example is at Vyachaslau Bortnik. Ref 1 is duplicate of refs 3, 4, 5, 6, 14, 18, and 19. Same situation with refs 2 and 7. Once again, the first diff contained several red-linked categories, which were later removed. Fortek67 (talk) 00:22, 7 June 2026 (UTC)

Sidenote – I've marked this request as resolved because everything seems to have now been done. Now we just have to wait the few days for all articles to be deleted. Whilst for the user's edits, they all seem to be the articles they created. Before that was just doing random stuff to get autoconfirmed granted. Fortek67 (talk) 00:26, 7 June 2026 (UTC)

This is tragic, because it appears to be done for Pride Month by a clueless person. Bearian (talk) 00:29, 7 June 2026 (UTC)

It's pretty unfortunate that a lot of articles I review now via AfC are AI generated... many of the reviewers don't even realise it and accept them. Then someone sees them months later and reports them.

Well at least the user isn't getting blocked for it – they still have the chance to actually write articles. Fortek67 (talk) 00:33, 7 June 2026 (UTC)

Random thought, is there a newsletter for AFC reviewers that could perhaps have a reminder on checking for AI-generated articles? In solidarity, Blue-Sonnet (I'm listening) 02:59, 7 June 2026 (UTC)

Greetings to all, as the discussion appear to be resolved, I would like to clarify that while I used some LLM assistant, sorry for that, the article was not entirely generated by LLM, it was primarily translated and adapted from another Wikipedia edition to English Wikipedia, if the articles does not meet English Wikipedia policies or guidelines. I understand that deletion may be appropriate. However, I am also willing to improve the article, address any concerns, or have it moved to draft space for additional work and sourcing. Thank you to everyone who participated in this discussion and review process. AMurmu (talk) 03:05, 7 June 2026 (UTC)

This comment itself reads as AI-generated or 'assisted', which does not help your case. Athanelar (talk) 07:23, 7 June 2026 (UTC)

ok AMurmu (talk) 07:53, 7 June 2026 (UTC)

It's been tried. It led to a lot of opposition to the very idea of AI cleanup and detection Gnomingstuff (talk) 14:56, 7 June 2026 (UTC)

A lot of opposition? Where can I find the thread? Fortek67 (talk) 15:30, 7 June 2026 (UTC)

somewhere on the AfC or NPP talk page, also featuring the charming dismissal of "every time something is flagged for AI I assume it is wrong" Gnomingstuff (talk) 22:25, 9 June 2026 (UTC)

It might be worth trying again. Public opinion seems to be shifting. M kuhner (talk) 03:55, 9 June 2026 (UTC)

 Zayyehtet212 (talk · contribs · logs · block log) (8 June 2026)
This matter has been resolved; no further action is required. 

@User:Zayyehtet212 has repeatedly been creating AI-generated pages, e. g. . Their behaviour also extends to spamming talk pages with AI summaries. Their repetitive editsums suggest that they are either an agent, or are blindly copy-pasting into the edit box. This user has been warned when one of their articles was draftified, but has not made any meaningful communication. Somepinkdude (talk) 18:11, 7 June 2026 (UTC)

This just looks like a human who edited normally in 2024, then switched to AI in 2025/2026. Sarsenet•he/they•(talk) 22:45, 7 June 2026 (UTC)

Not so sure about that; this reads as pure AI as well (from 2024). Somepinkdude (talk | contribs), in solidarity 16:18, 8 June 2026 (UTC)

Well spotted – though this seems to be the only one in 2024. Not sure what to do with that page though, maybe CSD it? Fortek67 (talk) 16:24, 8 June 2026 (UTC)

Yup, also looks like the user tried removing the LLMPROD, luckily it was reverted. I've went ahead with collapsing the LLM comment at Talk:Horonai Railway. Next time add {{aitop}} and {{aibottom}} if you see an LLM-generated comment. I've also posted a comment and tagged Draft:Maung Sein Win (Puti Kone) and Draft:Htoo Ant Lwin for AI. Not sure if it should be CSDed. All cleanup should be done, let us know if the user uses AI again. Fortek67 (talk) 16:19, 8 June 2026 (UTC)

CSD'd the user page with G11; unambiguous promotion is quite standard for pre-2025 LLMs. Somepinkdude (talk | contribs), in solidarity 16:46, 8 June 2026 (UTC)

 Elenareadwood (talk · contribs · logs · block log) (10 June 2026)
Editor has been blocked site-wide for AI/LLM policy violations. 

Y Cleanup has been completed.

CleanWorkplace impact of artificial intelligence  reverted edits

CleanData center  reverted edit, which was in wrong part of article

CleanLarge language model  reverted edit, which was in wrong part of article

CleanWorld Intellectual Property Organization  reverted by Asukite

CleanHugging Face  reverted promotional edit

CleanAI Overviews  reverted by Asukite

CleanArtificial Intelligence Act  reverted edits introducing source/text mismatch

CleanChatbot

CleanRetrieval-augmented generation

CleanArtificial intelligence in healthcare

CleanGenerative AI

I noticed that HJ Mitchell blocked Elenareadwood for spam, two accounts so far, possibly goes back further. I've listed the articles where their contributions have not yet been reverted as they may need checking. —In solidarity with Wiki Workers United · ClaudineChionh (she/her · talk · email) 23:21, 10 June 2026 (UTC)

@ClaudineChionh I meant "AI spam" as in they were spamming an AI site, I didn't look at whether their text was AI-generated but I reverted the spam. They're exactly the same as Claramilestone. If you have any suspicions about the integrity of their other additions after reviewing a sample, I'd suggest we probably won't lose anything of value if you reverted all of them before they're buried in page histories. HJ Mitchell | Penny for your thoughts? 10:03, 11 June 2026 (UTC)

I have now examined several (marked above). In one case Elenareadwood's edit introduced a source/text mismatch of a fairly LLM-like nature, though with only one short edit it's hard to tell (Artificial Intelligence Act). In one case it just made hash of a paragraph about a different topic (Workplace impact of artificial intelligence). In Hugging Face the edit was very promotional with some peacock wording: possible COI here? I am going to check and probably revert all the others if I can. M kuhner (talk) 02:22, 12 June 2026 (UTC)

(edit conflict)I cleaned the rest I've found here, citing LLMPROD. A few had already been gotten to, thankfully these were recent and easy to remove. Applying the mark reverted userscript to the user's contribs shows that they are all undone. Not sure about any other related users, however. ASUKITE 02:44, 12 June 2026 (UTC)

Done a bunch more. The editor's frequent modus operandi was to drop a chunk of text and a source into a random part of the article, as if they were more interested in getting that chunk published than in the logic flow. The sourcing is often correct but the overall effect is detrimental. The inserted material is also very pro-AI, every time. M kuhner (talk) 02:40, 12 June 2026 (UTC)

Sorry, I didn't realize you were actively working on this! I'm still kinda new to this process. ASUKITE 02:46, 12 June 2026 (UTC)

Not at all! It's very satisfying to see that between three of us, it's suddenly all done!

M kuhner (talk) 02:48, 12 June 2026 (UTC)

Yeah, I'm rereading LLMPROD and realizing it may have been a bit hasty, but if anybody here disagrees it should be easy for me to undo my removals. I almost feel like waiting a week is going to make removal more difficult. Reading HJ Mitchell's comment above, a narrow consensus might already be plausible, but next time I'll try to make sure it's more explicit. (I will however note that a CU block for multi-account abuse is unlikely to be overturned anytime soon, so perhaps it's okay) ASUKITE 02:53, 12 June 2026 (UTC)

I am not sure they were LLMPROD, but every one I looked at was bad: I had no qualms about reversion. I doubt you removed anything anyone will miss. I kind of feel the editor had a page of chunks in front of them and was just grabbing a chunk and dropping it randomly into the article. Whether the chunks were LLM-generated or not...in one case I'm pretty sure yes, in the others they're too short. But the dropping into random places is bad in itself. And--I never saw the original spam, but if Huggy Face isn't COI I'll eat my hat. M kuhner (talk) 03:01, 12 June 2026 (UTC)

Thanks @M kuhner and @Asukite: some of these came across my radar in recent changes and I'd flagged them to look at on Friday or Saturday, but you've taken them off my plate! —In solidarity with Wiki Workers United · ClaudineChionh (she/her · talk · email) 05:23, 12 June 2026 (UTC)

Y Cleanup has been completed.

Hi, I am having an issue with a TA (~2026-34262-43) repeatedly restoring a large LLM-written section into the article Microplastics and human health.

Sorry, I'm not sure what is the correct process with this. I gave them a WP:3RR warning on their talk page, but they just gave an LLM reply and restored the content anyway.

Thanks, CommonsKiwi (talk) 02:42, 11 June 2026 (UTC)

Now ~2026-33648-34 (the TA who originally added the material) is also restoring the content and using LLM-generated text in their edit summaries. CommonsKiwi (talk) 13:29, 11 June 2026 (UTC)

Both of the temp accounts you mention have been indeffed by ToBeFree. If any of the LLM material remains, you can remove it yourself or post here for us to do it if you feel you've done too many reverts already. If more temp accounts appear to try again, I suggest contacting ToBeFree directly, as blocking the accounts means we can avoid playing whack-a-mole with the content. Thanks for reporting this! M kuhner (talk) 03:16, 12 June 2026 (UTC)

ToBeFree also temporarily protected the page so it can't be edited by newly created temps. Hopefully that does the job. M kuhner (talk) 03:31, 12 June 2026 (UTC)

Thanks, I had seen that, but forgot to update my comment here. I don't think any further action is required, so I'll change the cleanup status to 'completed'. Cheers, CommonsKiwi (talk) 05:11, 12 June 2026 (UTC)

Y Cleanup has been completed.

Pretty cut and dry case: Talk:Operation Southern Spear has a TA making an AI-written move request. The request itself is facially inappropriate, the TA is disruptively making multiple !votes. One of the TA's in question by the OP has been blocked for AI usage (by me), such that WP:LLMPRV is satisfied and the content can be presumptively reverted. However, because I commented on the move discussion before realizing it was AI-written, it's inappropriate for me to execute the revert, so I'm notifying here. The AI usage is obvious -- you can see from the blocked TA's talk page in their responses to being warned about the Arab-Israeli conflict, that they are not familiar with English; however their AI-written posts suddenly have expert-level English language fluency. ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 17:00, 11 June 2026 (UTC)

I am sorry, I am not clear what edit you are asking us to revert. Could you spell it out a little more? M kuhner (talk) 21:45, 11 June 2026 (UTC)

All LLM-written edits by the TA, but I can deal with the rest; the ones I need help with specifically are those in the section "Title change to Cartel War...". Which in this case, means closing the requested move as inappropriately created, since several of those edits were opening that discussion, and there have been subsequent responses to it (per WP:HATGPT). ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 22:33, 11 June 2026 (UTC)

I have done a non-admin close on the discussion and hatted the offending material. I am somewhat uncomfortable deleting the entire discussion due to multiple responses by other editors. M kuhner (talk) 00:57, 12 June 2026 (UTC)

Thanks -- sorry that I'm not being very clear tonight, the close was all I was asking for (I was pointing out the subsequent responses as being a reason why it couldn't simply be cleanly deleted). ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 01:00, 12 June 2026 (UTC)

Y Cleanup has been completed, see /2026-06-12 SamyakSamantasinghar for reference.

SamyakSamantasinghar (talk · contribs)

fuck you. You whore mere pieces of shit blanked out my whole project just because i used ai to write despite i personally read all the documents and then prompted the ai to write even 3 times fact che... , ... i openlu accept i used to code but always fact checked what the ai presented before submitting in preview section. Fuck your age old rules. If you fucking guys have guts delete my account not only indeff ban. I say fucking delete. Dont worry about my other works all are ai written. , anyways, user is now unsurprisingly blocked for disruptive editing. Seeking consensus for presumptive removal. fifteen thousand two hundred twenty four (talk) 20:36, 11 June 2026 (UTC)

• Support presumptive removal, you could not ask for a more clearcut rationale than that. M kuhner (talk) 21:39, 11 June 2026 (UTC)

• Support. It'd be a waste of time to verify the content. They made large edits minutes apart, so I assume they did not actually verify anything and were just lying.

MeowsyCat99 (they/them) (meow) 21:48, 11 June 2026 (UTC)

Just want to point out that there is no exception to WP:NOLLM for "verified" content either. fifteen thousand two hundred twenty four (talk) 21:52, 11 June 2026 (UTC)

Support but I will take comments like this any day over the evasive ones Gnomingstuff (talk) 03:01, 12 June 2026 (UTC)

• Support - no obvious reason not to clean up.ChrysGalley (talk) 13:28, 12 June 2026 (UTC)

Presumptive removal applied, see tracking page. Nothing more to do. fifteen thousand two hundred twenty four (talk) 01:14, 13 June 2026 (UTC)

 Dahproman (talk · contribs · logs · block log) (7 May 2026)
This issue is currently under review. List only contains created articles, other edits need to be looked at and potentially added. InfernoHues (talk) 02:11, 8 May 2026 (UTC)

AfDZeeno Foster  InfernoHues  nom'd for LLMPROD

AfDTeeMirror  InfernoHues  nom'd for LLMPROD

AfDKing Kaly  InfernoHues  nom'd for LLMPROD

AfDTolex (singer)  InfernoHues  nom'd for LLMPROD

AfDTomi Folayan  InfernoHues  nom'd for LLMPROD

DeletedCarterson  InfernoHues  already deleted at AfD

AfDRestlezz  InfernoHues  nom'd for LLMPROD

Created the AISIGNS-ridden Draft:Jidex and has created numerous other articles that I think are therefore worth some scrutiny, but I can't be bothered to do so myself right now. Athanelar (talk) 13:35, 1 May 2026 (UTC)

In my opinion every new article they created has clear AISIGNS. —Em-as-in-emily (talk) 16:45, 1 May 2026 (UTC)

Specifically, there's a lot of undue emphasis on notability (e.g. "He gained prominence with" or "He gained recognition for") and promotional fluff ("Temiloluwa Bayode is the founder and chairman of the Reflectors’ Foundation, a non-profit organization that has carried out multiple community-oriented projects"; "He has conducted workshops and collaborated with organizations such as the South African Broadcasting Corporation (SABC), focusing on training and capacity-building for emerging sound engineers and musicians in Africa"—emphasis mine). —Em-as-in-emily (talk) 16:52, 1 May 2026 (UTC)

I went through the sources of the article TeeMirror and there were lots of verification issues indicating either LLM use or inexperienced editing. In addition to numerous citations that failed verification that I tagged, some citations are obviously incorrect: one citation is to a Wikipedia article, another to an image gallery, and a couple to the home pages of websites instead of a particular page. Reading through the cited "news" articles, some of them also read as LLM-generated to me. I don't think the article is WP:G15-worthy but it reads to me like it was written using an LLM to find or summarize sources.

After this I am almost certain that all of these new articles on musical artists are worthy of deletion, but as I'm still new to this project I would appreciate a second opinion. —Em-as-in-emily (talk) 17:19, 1 May 2026 (UTC)

Please @Em-as-in-emily expand on this i need to know what LLM use issue is with the article on TeeMirror and the fix you suggest, because if we are dealing with the issue of AI generated content you haven't explained which part of the article is actually AI generated, you said "Reading through the cited "news" articles, some of them also read as LLM-generated to me." how can that be brought to this notice board if it's a matter of how the news article was written by the journalist of the News Body that published that article and i think every news reference used in TeeMirror also follows Wikipedia:WikiProject Nigeria/Nigerian sources. This article has already been reviewed and is only here in suspicion of having LLM content, as a community we have to work together to fix an issue, so kindly explain which specific text/paragraph not citation, not webpages or wikilinks, but a paragraph or anything like that and i will personally make the research and fix it. Dahproman (talk) 15:20, 11 May 2026 (UTC)

Having received a series of advice from InfernoHues, I will ensure to take responsibility if any of these suspected articles have AI content. I would say here, just for the record, that these articles have already been reviewed before now. Also, I do not write or use AI to generate content for the articles I have written. I started using AI in late 2025 for things like punctuation and concise spell checks, which WP:LLM supports; but for the article Restlezz, which was written in mid-2024, we were all still afraid of ChatGPT at the time, so there was no AI whatsoever there.

I never want to, nor would I, use AI LLMs to generate content because that is not ethical in any sense whatsoever. As a software developer myself, seeing the dangers that LLMs bring to the hands of the general public, I totally understand the need for everything this noticeboard stands for. On that ground, I will closely review and rewrite any sections that have been flagged by you guys as potentially AI-generated to ensure the tone is neutral and meets this community's standards.

Also, guys, I will point out an issue that is usually overlooked in the Wikipedia space as well: editors making contributions to an article without understanding the issues that writing with LLM-generated content brings. Just for the feel of a quick edit to improve their edit count, they rush using LLM chatbots to get random content and just make the edit; in the end, the creator of the article is the one who bears the burden or, in cases like this, appears on a noticeboard. That is why I said instead of these articles being deleted, let each scenario of AI-generated content in that article be identified and resolved with suggestions made instead, When editing a Wikipedia article for the first time, you see this notice:

"Notice about sources This article must adhere to the biographies of living persons policy. For all related material: Take extra care to cite high-quality sources when adding or changing information. Do not add content cited only to tabloid journalism. Do not use self-published sources unless written or published by the subject. Immediately remove contentious material that is uncited or cited to low-quality sources. If inappropriate material is repeatedly inserted, or if you have other concerns, please report the issue. If you are connected to a subject of this article, help resources are available."

I think to help buttress this point, a new point should be added to this notice like: "When editing with LLMs, make sure to adhere to WP:LLM before placing edits." Something like that, guys, is my idea; might be foolish, but could be something. Dahproman (talk) 14:58, 11 May 2026 (UTC)

In your Wikidata unblock request, you say you work with Search Alpha, an AI SEO marketing service. I'm concerned that these articles were created for promotional purposes. InfernoHues (talk) 15:50, 11 May 2026 (UTC)

I wouldn’t say work for Search Alpha, but more like collaborated with them for a brief period of time on a project. Well, let me just put it out there: Search Alpha has been working with Google Search on some projects, and they were trying to work on a Wiki tool that helps SEO SaaS companies integrate Wikipedia easily into their infrastructure in a way that they can utilize Wikipedia’s MediaWiki API to create tools to improve search tools that are provided and verified solely with the MediaWiki API. But the project is abandoned right now, so no, not with Search Alpha just saw a project that needed an expertise, it wasn’t as possible as we thought, then I left, guys in 2026 the dependence of wikipedia as the source of truth is waning due to chatbots and a few other sources discrediting wikipedia we thought finding a way to get SAAS tools to integrate wikipedia as a source of dictionary to the world would be effective in keeping wikipedia at the top. Dahproman (talk) 16:39, 11 May 2026 (UTC)

I've reinstated the tags you've removed from articles without addressing the issues in said articles.

Please do not do that again. Blue-Sonnet 05:57, 15 May 2026 (UTC)

Hi @Blue-Sonnet, I'm new to Wikipedia:LLMPROD , so I need a few pointers. I had no idea that there was anything more to do. There has been only one article discussed, that's TeeMirror, and there was nothing specific i think, that was discussed that needed addressing; I had asked for a direction on what to do. I haven't seen anything so far to update. Then, for the remaining articles that I removed their tags, the former tags had said if the tags weren't removed the articles would be deleted at a said time; and seeing that no issues at all have been raised to be addressed for those other articles, I removed the tags.

Look, I am ready to do the work, but so far so good, for the 6 articles mentioned in this noticeboard only one has been discussed. That was why I removed the tags I need to see a reason so I can fix it, i am new to this AI noticeboard, instead of this going on and on without any solution which could lead to the deletion of the article that could be avoided. So guys, help me out; give the reasons each of these articles was nominated for WP:LLMPROD and let's work together to fix it. As i said I'm new to WP:LLMPROD this is my first AI noticeboard. cheers Dahproman (talk) 19:25, 15 May 2026 (UTC)

@Dahproman Take a look at WP:LLMPRVOBJ. If you take issue with an LLMPROD tag, you're expected to make an effort to fix the issues in the article before you remove the tag. Simply removing the tag because you don't think it was suitably justified is not appropriate.

The way PROD works is that the admin who deletes the article after those five days first makes sure that the tag is actually appropriate: so if there's no good reason for the tag then the deletion will be denied at that stage. Athanelar (talk) 19:45, 15 May 2026 (UTC)

The TeeMirror concerns others brought up are just examples of LLM text. You should find the other LLM text in the other articles and TeeMirror and fix it. InfernoHues (talk) 20:01, 15 May 2026 (UTC)

As part of the prod objection process, the objector must ensure the content meets all polices and guidelines first, this includes WP:NOLLM. So if an editor has added generated text, they must remove, not correct, it first before deprodding. fifteen thousand two hundred twenty four (talk) 20:07, 15 May 2026 (UTC)

ok i will do that now thanks for clarifying this. @InfernoHues Dahproman (talk) 20:53, 15 May 2026 (UTC)

Hi @InfernoHues all of the articles have been deleted and moved to draft before enough time to work on them properly so i will use this time to properly fix the draft or all any thing that might go against Wikipedia:Presumptive removal of AI-generated content and then fix it before moving to the main space again. Dahproman (talk) 17:59, 17 May 2026 (UTC)

It might be best to submit the cleaned up draft (as well as any newly written ones) through the Wikipedia:Articles for creation process; that way an independent editor can double-check everything and make sure there won't be any further issues - they can then move the article on your behalf once everyone is happy with the result. Blue-Sonnet 18:06, 17 May 2026 (UTC)

Hi @Blue-Sonnet i think there should be a way to get the creators of articles to get enough time to review the articles and fix them while they can, i had 7 articles nominated with only one actually suspected of anything then all of them were deleted it doesn't sound right to me, i found this excess, i will work on this just as you've said and fix them but just for the record this is what i feel. Dahproman (talk) 18:15, 17 May 2026 (UTC)

 Rizwan (talk · contribs · logs · block log) (17 May 2026)
This issue is currently under review. 

 Cleanup has been requested.

Enormous amounts of AI edits. They have been warned about AI use before and did not respond. Gnomingstuff (talk) 20:49, 17 May 2026 (UTC)

@Rizwan: Hi Rizwan, can you please respond to this? LLM-generated content is prohibited on Wikipedia Kowal2701 (talk, contribs) 22:46, 17 May 2026 (UTC)

I usually write in my native language and use ChatGPT only to translate my writing into English. I never use AI to generate entire texts and then copy-paste them. Every single line had proper references provided, and not a single statement was made up. Everything was fully source-based. Yet you removed all the information instantly.

I write the content myself in Bengali, use AI only for translation, then carefully review it again before adding it. And it was nighttime here for us, so naturally it would take some time for me to respond. Rizwan (talk) 02:45, 18 May 2026 (UTC)

The writing was clearly enough ChatGPT that it was identifiable purely by reading it, which suggests it may not be a very faithful translation. Gnomingstuff (talk) 02:58, 18 May 2026 (UTC)

So how am I supposed to do translations then? I already explained that I write in my native language first and then use translation assistance. Not a single piece of writing was AI-generated. Maybe it looked that way to you because it was translated by ai. And was there even a single line without sources? Rizwan (talk) 03:35, 18 May 2026 (UTC)

Editors of the English Wikipedia are expected to be able write in English, if you must use an LLM to translate your writing into English then this is not the project for you. fifteen thousand two hundred twenty four (talk) 03:40, 18 May 2026 (UTC)

Listen, I do know English. But I use AI to translate my writing and then review it myself. Also, could you please give me the link to that Wikipedia policy? Rizwan (talk) 03:52, 18 May 2026 (UTC)

See the below comment. fifteen thousand two hundred twenty four (talk) 03:55, 18 May 2026 (UTC)

There are Wikipedias in languages other than English. Gnomingstuff (talk) 06:08, 18 May 2026 (UTC)

The same way that translation have always been done; by using your knowledge and understanding of the source language and the target langue and writing a faithful and good translation yourself. Translating any body of work is a work-intensive task, and often more difficult than writing the original text. If you lack the competency to write in English in the first place, then you have no business translating anything into English. If you cannot write your responses and contributions to English Wikipedia in English, then you do not have the required competency to contribute here, and should instead contribute to a different Wikipedia in a language that you do have the required competency to contribute in. --Gurkubondinn 11:32, 18 May 2026 (UTC)

That described usage does not meet either of the two exceptions to WP:NOLLM, it is neither basic copyediting, nor does it meet WP:LLM-assisted translation. Please stop. fifteen thousand two hundred twenty four (talk) 03:02, 18 May 2026 (UTC)

So much well-sourced information was removed all at once. What’s supposed to happen to all of that now? At the very least, a notice could have been placed at the top of the article first. Removing everything instantly like that doesn’t really seem right, especially when all the information was properly sourced. Rizwan (talk) 04:23, 18 May 2026 (UTC)

Some quick history may help things make sense.

LLMs generate highly asymmetric workloads for volunteers, they are very good at generating highly plausible, but ultimately flawed output. Very (very) often the burden of cleaning up this flawed output fell not on those who added it, but on the rest of the community. An editor would generate plausible junk, then other editors would have to spend even longer trying to fix it, this harms the project. And so the project adopted WP:G15, then WP:NEWLLM, then WP:NOLLM in order to protect itself. fifteen thousand two hundred twenty four (talk) 04:55, 18 May 2026 (UTC)

All the well-sourced, properly referenced lines of information have been removed! So what’s the solution to this now? Rizwan (talk) 06:09, 18 May 2026 (UTC)

Rewrite it in English without AI. That is the only solution. Gnomingstuff (talk) 07:31, 18 May 2026 (UTC)

Write it in English from the start. Let yourself look up words or phrases, but don't translate anything long. SenshiSun (talk) 14:37, 18 May 2026 (UTC)

 Cleanup has been requested.

Multiple claims failing verification, sentence fragments and some dramatic phrasing ("very daunting task," "fierce campaign," "huge refurbishment") Spelling/grammar errors make this look like ai-assisted writing. Other pages by same author may have similar issues. I left a note on their talk page (User talk:Glowr Curt Ryder) to which they claimed to have used AI for sourcing and verify "a few things", however the issues verifying sources (by myself and another reviewer) give some doubts. ASUKITE 16:58, 17 May 2026 (UTC)

• Draft:Garth, Gwynedd (edit | talk | history | links | watch | logs)

The claim that it would not be until 1922 is sourced to a source published in 1914, so this should technically meet the G15 criteria. --Gurkubondinn 17:06, 17 May 2026 (UTC)

Recent LLM writing has flawless grammar and spelling, and the writing in this article does not strike me as plausibly LLM.

However, there are very serious sourcing issues. I spot-checked sources and about half of them appeared to be wrong (the claim made in the article is not confirmed by the source). Some are also highly unsuitable sources.

One example: The chapel was built in the arts and crafts style. It had was built with an imposing stone facade, intricate window archery and a unique nonconformist design. This is sourced to which is a real estate ad for one of the apartments that have been build inside the former chapel. The real estate ad contains a link to a real estate brochure which does say that the chapel was build in Arts and Crafts Gothic style, but does not say anything about its facade, windows, or nonconformist design. So this is a source/text mismatch, and also an unreliable source.

Another: This kept the BBC and its radio on the air, before and after the blitz had occurred in London. is sourced to which is about a chapel of a different name, on a different street, built in a different year--surely not the same building as the rest of the paragraph.

It's not acceptable to use LLM to find sources unless you then read them, carefully, yourself (LLM summaries cannot be trusted) and make sure they are reliable sources and support the claims being made. That clearly did not happen here.

M kuhner (talk) 03:54, 18 May 2026 (UTC)

The article Bryntwrog by the same editor again has sourcing almost solely from rental listings for a cottage located there. Literally nothing about the hamlet itself. M kuhner (talk) 04:16, 18 May 2026 (UTC)

I have proposed Bryntwrog for deletion. M kuhner (talk) 14:43, 18 May 2026 (UTC)

You might want to read § Your reversion of my Bangor edit on my talk page. --Gurkubondinn 14:51, 18 May 2026 (UTC)

Thanks for bringing that to the discussion. Do you think the editor will now refrain from using LLM sourcing without checking it? I'm pretty uneasy about their proposal to use AfC as a correctness check; they really need to be doing some correctness checks of their own. M kuhner (talk) 15:12, 18 May 2026 (UTC)

Using AfC as a "correctness check" aside, I am happy to AGF and assume that Glowr Curt Ryder will refrain from using AI and has understood why NOLLM exists. I don't imagine that there are many editors with an interest in Welsh history, so hopefully they will prove me right and contribute positively to the project. --Gurkubondinn 16:05, 18 May 2026 (UTC)

I very much hope you're right. You may want to comment on the deletion discussion: Wikipedia:Articles_for_deletion/Bryntwrog. (Unfortunately I am not convinced this can pass notability, LLM issues aside.) M kuhner (talk) 16:23, 18 May 2026 (UTC)

Maybe, I'm still considering it. I tend to lean inclusionist, and my participation at AfD has mainly been arguing on the basis of NOLLM reasons. I'm mostly worried about WP:DONTBITE here. --Gurkubondinn 16:32, 18 May 2026 (UTC)

I take your point and will try to be more patient in the future. M kuhner (talk) 16:49, 18 May 2026 (UTC)

No worries, it's something that I have had to work on myself. --Gurkubondinn 17:05, 18 May 2026 (UTC)

There is also WP:TEAHOUSE § New here, where Athanelar gave a masterclass in firmly explaining NOLLM without biting the newbie. --Gurkubondinn 17:17, 18 May 2026 (UTC)

 PeppermintSA (talk · contribs · logs · block log) (13 May 2026)
Editor has been warned about the use of AI/LLM content and WP:NOLLM. 

Needs reviewFK Sarajevo records and statistics

Needs reviewKK Bosna Royal

Needs reviewBoris Jordanovski

Needs reviewMartin Vanjak

Needs reviewFaruk Kulović

Needs reviewAleksandar Avlijaš

Needs reviewDamir Abdulkerim Mirković

Needs reviewAdmir Bukva

Needs reviewVojin Ilić

Needs reviewAlen Abaz

Needs reviewSabahudin Bašović

Needs reviewFilip Živković

Needs reviewRijad Telalović

Needs reviewSerhiy Ihnatkov

Needs reviewDamir Zeljković

Needs reviewMirsad Milavić

Needs reviewSenad Redžić

Needs reviewMilenko Novaković

Needs reviewMiroljub Mitrović

Needs reviewJosip Pandža

Needs reviewIbrahim Krehić

Needs reviewNovak Manojlović

Needs reviewDraško Prodanović

Needs reviewZdravko Čečur

Needs reviewMilivoje Karalejić

Needs reviewEmir Halimić

Needs reviewHamdo Frljak

Needs reviewDženan Rahimić

Needs reviewClarence Gilbert

 Cleanup has been requested.

• PeppermintSA (talk · contribs · deleted contribs · logs · filter log · block user · block log)

Extremely prolific AI use (disclosed as ChatGPT on their talkpage) going back a long time -- this user is singlehandedly responsible for "Sarajevo"-related tokens showing up disproportionately in my datasets, because of the sheer amount of their drafts that were rejected for LLM usage -- and still going to this day. They said they'd stopped using AI, but articles like this make that hard to believe. Gnomingstuff (talk) 16:24, 13 May 2026 (UTC)

Not sure why this is marked as resolved, this person has dozens of articles, not 1 Gnomingstuff (talk) 06:33, 15 May 2026 (UTC)

Yeah this and the previous editor have so many contributions that we have to dig through to even compile a list. ClaudineChionh (she/her · talk · email · global) 03:21, 19 May 2026 (UTC)

Oh, I see ... the "article cleanup list" comes from the beta version of the AINB Helper script. ClaudineChionh (she/her · talk · email · global) 03:58, 19 May 2026 (UTC)

Bmgupta12345 (talk · contribs · deleted contribs · blacklist hits · Edit filter search · what links to user page · count · COIBot · Spamcheck · blacklist hits · user page logs · x-wiki · status · Google · StopForumSpam)

Some stuff to clean up here. If anyone else could contribute it would be great. Thanks, CommonsKiwi (talk) 02:39, 20 May 2026 (UTC)

Are all of their edits LLM-assisted, or did this start after a certain date? A little bit more info about this could help. Sarsenet•he/they•(talk) 02:44, 20 May 2026 (UTC)

Not sure, they seemed to start by doing a lot of small fixes/copyediting. From what I saw these edits looked fine, even if they were AI-assisted. What concerns me is where they have added content, which has become more frequent over time. CommonsKiwi (talk) 02:54, 20 May 2026 (UTC)

The small edits are absolutely AI generated (note how the edit summaries get cut off, suggesting they didn't even bother to check what they pasted in). And from the looks of it, it is yet another deluge of Newcomer Tasks-induced AI editing. One day WMF will stop destroying their product with enshittification features. Gnomingstuff (talk) 03:54, 20 May 2026 (UTC)

User page also looks like it was done by ChatGPT. Sarsenet•he/they•(talk) 06:12, 20 May 2026 (UTC)

This user was already warned on May 14 to stop creating articles using LLM tools. Instead of responding to the message, they have continued creating AI-generated articles. Two days later, they created Peace Effiong, which includes the sentence: She has been listed among the attacking players representing Rivers Angels in youth competitions and development tournaments connected to Nigeria’s national team scouting system Just today they created Yusuf Abdullahi Tsamage, which says He has been described in match reports as an attacking player deployed in advanced positions This is clearly not verified by the attached source. JTtheOG (talk) 03:58, 21 May 2026 (UTC)

Seemingly AI-generated revision to Bocconi University School of Law; uses primary sources mainly and references its own Wikipedia page. Subsequent discussion on user talk page with possible sock/meatpuppet seems to indicate that they may be using LLM-gen text to respond to comments as well. It’s very late where I am at the moment so I don’t have the energy to deal with this. If someone who’s skilled in dealing with things along these lines can step in, I’d appreciate it. OrbitalVoid49 (talk) 11:25, 20 May 2026 (UTC)

I inform you that there is no AI response system, I believe that these accusations are falling into defamation. I implemented a page that lacked a series of information by putting clear, external sources (such as the University website) in addition to other sources. I believe that on your part there is only the will to mock the work of others instead of improving Wikipedia.

In addition, I work with my name and surname not with a false name aimed only at protecting yourself ~2026-30092-25 (talk) 11:29, 20 May 2026 (UTC)

So is the ~2026 your surname, or is it the 30092-25?

I don't know what an "AI response system" is but that's not what they're asking. They are asking whether you (assuming you just aren't logged in currently) used AI to write the article, which is a question that has been dodged over and over here. Gnomingstuff (talk) 12:31, 20 May 2026 (UTC)

Even if it's untrue, "you used AI for editing Wikipedia" is not libel. Hint: more people sensed that you have used AI. But since it is not a crime, nor a reason for social dishonor, it does not amount to libel. I can say "you did not brush your teeth today". It is a lie, but not libel. tgeorgescu (talk) 22:17, 8 June 2026 (UTC)

Disruption continues. More eyes needed. tgeorgescu (talk) 19:26, 10 June 2026 (UTC)

Considering the article got turned into a redirect and the user itself got blocked from editing the page, should this be considered closed? OrbitalVoid49 (talk) 21:31, 13 June 2026 (UTC)

Y Cleanup has been completed.

• ChristinePgh (talk · contribs · deleted contribs · logs · filter log · block user · block log)

Found this user while searching for placeholder URLs in references: Diff/1339147116. Lots of large rapid-fire edits and page creations. ‑‑gurkubondinn 13:13, 16 June 2026 (UTC)

has markup Kowal2701 (talk, contribs) 14:55, 16 June 2026 (UTC)

And why does the edit summary say [e]xpanded stub? This is the first revision of a new page.. ‑‑gurkubondinn 15:00, 16 June 2026 (UTC)

not telling you anything you don't already know but the edit summaries are just pure chatbot responses. I like this one:

Important Strategic Insight This letter allows you to: Connect Fitzhugh to early frontier defense Show pre-war tension Strengthen your War of 1812 narrative Elevate Sodus beyond a minor settlement Gnomingstuff (talk) 22:54, 16 June 2026 (UTC)

I think that I've cleaned up everything that I could after this user, tagged the rest. Also obvious chatbot edit summary: Diff/1339354638 ‑‑gurkubondinn 09:57, 17 June 2026 (UTC)

Y Cleanup has been completed.

The Technological literacy page seems to have gotten a LLM-generated edit by a drive-by user here. ~2026-35622-93 (talk) 23:58, 17 June 2026 (UTC)

Thanks for flagging. Since the edit was recent and very little happened in the article since, I just reverted to the issue prior. Gnomingstuff (talk) 05:02, 18 June 2026 (UTC)

I have given the user a warning. Since this was their only edit, I am marking cleanup as completed. M kuhner (talk) 06:07, 18 June 2026 (UTC)