Wikipedia talk:Good article nominations
From Wikipedia, the free encyclopedia
This is the discussion page for good article nominations (GAN), good article reassessment (GAR), and the good articles process in general. To ask a question or start a discussion about the good article nomination process, click the Add topic link above. Please check and see if your question may already be answered; click the link to the FAQ above or search the archives below. If you are here to discuss concerns with a specific review, please consider discussing things with the reviewer first before posting here.
 | To help centralize discussions and keep related topics together, several other GA talk pages redirect here. |
| Main | Criteria | Instructions | Nominations | FAQ | Discussion | Reassessment | Report |
| Reviewing initiatives: | February backlog drive | Mentorship | Review circles | Pledges |
Proposal - add alt-text of images to GACR #6b
I believe that the Good Article criteria section 6b should be modified to include the use of alt text. This is part of the Manual of Style and is an accepted part of the FA criteria (as an extension of Wikipedia:Manual of Style/Captions). It is relatively simple to implement (there is and is a benefit for users with screen readers or those who cannot load images due to slow internet connections or other restrictions (ex: an internet block on Commons). Because "suitable captions" has always been part of section 6, this could be interpreted as always having been a part of the criteria, though it wasn't linked directly as it is in WP:FACR.
Current text: 6b: media are relevant to the topic, and have suitable captions.
Proposed text: 6b: media are relevant to the topic, have suitable captions, and include alt text.
The Manual of Style is already alluded to in 1b, so this isn't coming out of nowhere to mention this part of it. If this were to be implemented, it would be certainly be a project to go over the backlog of older non-compliant GAs and bring them up to the standard, but compared to more major issues that plague old GAs (unsourced passages?) it's much easier to fix. I do not see this as being an issue that would warrant a Good Article Reassessment on its own for this reason either. -- Reconrabbit 17:55, 27 February 2026 (UTC)
- I'm weary about adding even tighter expectations into the GA standard, especially when such standards already exist if wanting to get the article to featured status. If we add a little FA-criteria here, and a little there, the GA standard starts to become FA-lite. My view is that captions should be (and are stipulated as being) expected to accompony images, while an article isn't going to become a firmer "good" by adding alt-text to images. A reviewer may encourage it, but do we really need to mandate it for GA standard? Bungle (talk • contribs) 18:10, 27 February 2026 (UTC)
- I would support it, personally. It's not a big difference for the article creator - not much more work than adding the initial caption is in the first place - and is a big help for accessibility. I don't at all think it's WP:CREEP. -- asilvering (talk) 18:20, 27 February 2026 (UTC)
- I feel this is a bit of creep of rules. Are we really going to start to fail articles because there is no alt text? Lee Vilenski (talk • contribs) 18:52, 27 February 2026 (UTC)
- I certainly hope not. That would be an incredibly lazy fail, since it would take the reviewer only moments to write up some of their own if that were the only barrier to GA status. -- asilvering (talk) 19:00, 27 February 2026 (UTC)
- Quite - so why have it in the rules. I'd support a more general "article should meet the majority of the MOS". If someone says your images should have Alttext in a review, you just put it in. Why have a rule for it. Lee Vilenski (talk • contribs) 19:06, 27 February 2026 (UTC)
- But "should meet the majority of the MOS" is a huge expansion over the GA criteria! -- asilvering (talk) 19:11, 27 February 2026 (UTC)
- No, I agree. The wording I have is bad, but I don't really support very minor additions being part of the criteria. If we are to make the criteria larger, I'd want it to be for more general adherance the MOS, rather than one or many small parts of it. Lee Vilenski (talk • contribs) 19:16, 27 February 2026 (UTC)
- It would be on the same level as someone saying "your images don't have captions, the GA criteria says they should". It is a straightforward fix. -- Reconrabbit 19:16, 27 February 2026 (UTC)
- Seeing your second reply that came in right as I wrote mine - I see the issue that would come up from this exactly because it is such a small change. I just feel it is a simple step to make articles more accessible; putting it in as part of these criteria could make it more visible over the whole project as well. -- Reconrabbit 19:19, 27 February 2026 (UTC)
- But "should meet the majority of the MOS" is a huge expansion over the GA criteria! -- asilvering (talk) 19:11, 27 February 2026 (UTC)
- Quite - so why have it in the rules. I'd support a more general "article should meet the majority of the MOS". If someone says your images should have Alttext in a review, you just put it in. Why have a rule for it. Lee Vilenski (talk • contribs) 19:06, 27 February 2026 (UTC)
- I certainly hope not. That would be an incredibly lazy fail, since it would take the reviewer only moments to write up some of their own if that were the only barrier to GA status. -- asilvering (talk) 19:00, 27 February 2026 (UTC)
- Weak oppose for now per Kusma; although I agree that alt text should become standard for quality articles, current on-project guidelines for alt text (MOS:ALT and Help:Alt text) are too dubious to work with at the moment—they both contradict sources like e.g. section508.gov's advice, the former contradicts itself constantly on the level of detail to be used, and the MOS:BLANKALT advice on
|alt=icon/|alt=photograph/ etc. for decorative non-PD images is sketchy (perhaps even its definition of decorative images—maybe images that provide visual structure aren't decorative? I tried to contact section508.gov for advice, but my email didn't reach them). I think these guidelines need to be improved before we can make alt text something to be done systematically. {{Lemondoge|Talk|Contributions}} 22:39, 16 March 2026 (UTC)- Very good points, I often find myself trying to recommend simpler alt texts for images but struggling to find concrete guidelines on what is appropriate for a given figure. And that's just in the context of biology articles (when is it more appropriate to provide every aspect of a detailed diagram for alt text over just "diagram of a flower"?). -- Reconrabbit 13:20, 17 March 2026 (UTC)
Our GA rules deliberately include only some of the MOS and I'm not sure either way about whether we should incorporate all of Wikipedia:Manual of Style/Captions (which does include alt-text). However, I have a broader concern, that the more we creep our rules towards mechanically checkable things like the presence of footnotes and alt-text in certain positions, the more we are also encouraging reviewers and article maintainers to pass over the other aspects of our rules that are not so mechanical: are these references, images, and content the right references, images, and content for an encyclopedia article on this topic? For that matter, checking whether alt text exists is far from the same as checking whether it is good as alt text. —David Eppstein (talk) 20:37, 27 February 2026 (UTC)
- Generally, having reasonable ALT text would be a plus, but Wikipedia:Manual of Style/Accessibility/Alternative text for images is not a usable guideline that tells us how to write good ALT text and may even mislead us (see the discussion on the talk page). I do not think "images have ALT text, but it is not useful" is a big step up from "no ALT text", so I oppose for the moment. —Kusma (talk) 18:16, 28 February 2026 (UTC)
- A related question: supposing I want to make sure an article has appropriate alt text. Is there an easy way to do so, without groveling through the source code view of the article and without adding third-party extensions to my browser (Firefox)? My default user interface does not appear to show alt text. Going to the Firefox accessibility property inspector separately for each image works but is cumbersome and non-obvious. —David Eppstein (talk) 19:30, 28 February 2026 (UTC)
- One of the features of Wikipedia:Tools/Navigation popups is to show the ALT text for images on hover. —Kusma (talk) 06:59, 1 March 2026 (UTC)
- Ok, it seems you also have to disable "Redirect image links to Commons for files hosted there" (which I use much more frequently than wanting to find out about alt text) for this to work. Otherwise it does nothing on image hover. And after disabling commons redirects, as well as getting uglier popups on wikilinks, I get a popup on an image that shows me what its commons landing page would look like except that the templates are unexpanded. Because it's entirely about the commons landing page, I see nothing about its alt text here. Am I doing something wrong or is that what the gadget is supposed to do? —David Eppstein (talk) 08:14, 1 March 2026 (UTC)
- One of the features of Wikipedia:Tools/Navigation popups is to show the ALT text for images on hover. —Kusma (talk) 06:59, 1 March 2026 (UTC)
- A related question: supposing I want to make sure an article has appropriate alt text. Is there an easy way to do so, without groveling through the source code view of the article and without adding third-party extensions to my browser (Firefox)? My default user interface does not appear to show alt text. Going to the Firefox accessibility property inspector separately for each image works but is cumbersome and non-obvious. —David Eppstein (talk) 19:30, 28 February 2026 (UTC)
- This is getting into WP:instructions creep territory. We already have extra hoops with the new mandatory source review, and it hasn't exactly helped with the backlog, more the contrary. GAN risks becoming a single person FAC review (where alt-text isn't exactly required, just encouraged), which would be fine if we had a corresponding reviewer pool to pull from, but we don't. FunkMonk (talk) 17:27, 2 March 2026 (UTC)
- Weak support. I don't think this is instruction creep; it takes like less than 5 minutes to describe an image so I wouldn't mind I guess (I also think that failing articles solely because they don't contain alt text is incredibly lazy... it's like failing an article for failing to meet Criterion 1b because its lead violates MOS:FIRST despite the rest of it being fine). However, I am concerned that if we add it as part of the GAN criteria, then it would be a lot of work to add alt texts to existing GAs. Icepinner (Come to Hakurei Shrine!) 09:18, 6 March 2026 (UTC)
- Oppose; however, alt text can be made encouraged without being made mandatory (like advising that a good article candidate with un- or sloppily formatted citations have the citations formatted correctly for a future FA candidacy). I see that at least two other users have expressed concerns of instruction creep, so that's that. Whyiseverythingalreadyused (t · c · he/him) 02:18, 10 March 2026 (UTC)
- Weak oppose I think adding alt text is always best practice, but GA is already strict enough as is and I can't imagine I'd ever support delisting a GA based on missing alt text alone. Trainsandotherthings (talk) 21:05, 17 March 2026 (UTC)
AI concerns
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
Hello, I am reviewing an article & am getting bogged down in thinking it contains some WP:AISIGNS. I'm honestly at a loss for what action to take. I've posted my thoughts in the review. I can keep pointing out sentences that seem robotic or flowery to me, but that is ultimately a time sink for both me and the nominating editor.
Should I ask for a second opinion for this GA review? Do you all have any other advice for me? Thank you for your time. As I am a new editor and this is my first GA review, I want to get this right. Chao Garden 🌱 (hi) 07:01, 6 March 2026 (UTC)
- It'd be easier to give advice if you linked to the article in question. Mr rnddude (talk) 08:41, 6 March 2026 (UTC)
- Disney's Aladdin (Sega Genesis video game) Chao Garden 🌱 (hi) 08:44, 6 March 2026 (UTC)
- As a comment, I note that the nominator of the article is Cat's Tuxedo, an editor who has 74 GAs, 2 FLs, and 3 GTs. With a track record like this, I generally expect an editor like them not to use LLMs for article writing, though there could be a chance (not taking any sides here; I'm neutral on this, just pointing this detail out for further consideration). Icepinner (Come to Hakurei Shrine!) 09:10, 6 March 2026 (UTC)
- Ok. So, to start Cat's Tuxedo first edit to the article was Dec, 2018 and first significant expansion to the article in Dec 2019 – Some of the passages you flagged as AI concerns are from content written in this edit, for example:
Aladdin's primary forms of offense against enemy characters are a scimitar for short-range slashing attacks and apples that can be pelted as long-range ammunition. The apples are a finite resource, but can be collected in abundant amounts throughout the game
. This is well before the ChatGPT (or other AI) era (post Nov, 2022) and could not have been generated using it. Something just shy of half the article already existed pre-AI era (the article is ~27,000 bytes in Oct, 2022). This doesn't preclude the possibility of AI involvement, but it does indicate that you have false positives in your review. This is a recent (Feb, 2026) and quite substantial edit. There are passages within it that raise my eyebrow, for example[t]he game's development represented Disney's inaugural direct participation in video game production, motivated by a desire to produce a high-quality adaptation that matched the film's artistic and narrative excellence
is PEACOCK at the very least, but that still could have been human written. Mr rnddude (talk) 09:27, 6 March 2026 (UTC) Addendum: Separate question: Why do your comments have a mixture of quote mark styles? Mr rnddude (talk) 09:35, 6 March 2026 (UTC)- Do you think I'm over-analyzing this?
- Also my comments probably have a mixture because I wrote them in Apple Notes app and copied them over to format. Chao Garden 🌱 (hi) 09:41, 6 March 2026 (UTC), By the way, thank you for pointing out the details about the edit history. That's a good point & I didn't check it carefully before my last review comment that pulled out specific sentences.Chao Garden 🌱 (hi) 09:55, 6 March 2026 (UTC)
- Potentially. There are passages in the prose that raise my eyebrows and I am also at least a little wary of large individual edits, but these are spaced apart by multiple hours indicating an editor actively working on it before posting. For example on 14 Feb 2026 we have edits at 01:51 (+2,855 bytes), 05:37 (+3,794 bytes) and, 08:53 (+5,567 bytes). It is not unreasonable to be able to build any of those edits over the course of a couple hours and that is an indicator of human-made content. I'm inclined to give benefit of the doubt. I can also ping an editor that I have crossed paths with that works heavily in dealing with AI generated content on Wikipedia if they're willing to give their considered opinion if that'd be helpful. Mr rnddude (talk) 10:49, 6 March 2026 (UTC)
- @Mr rnddude I really appreciate your time. Would you be able to ping that editor you mentioned? I would love to know their take on this as well.
- Everything you've pointed out makes a lot of sense to me... the edit history is very strong evidence here. I was so wrapped up in the way things sounded and comparing it to WP:AISIGNS that I didn't pay as much attention to the history as I should have. So thank you again for breaking that down for me. In the future I'll definitely start there before nitpicking individual lines (especially lines that were clearly written pre ChatGPT). Chao Garden 🌱 (hi) 16:48, 6 March 2026 (UTC)
- Sure, no problem. Gnomingstuff – Your knowledge in identifying likely AI-generated content has been requested, if you wouldn't mind taking a look at Disney's Aladdin (Sega Genesis video game) and giving your thoughts. Mr rnddude (talk) 16:59, 6 March 2026 (UTC)
- No problem, sorry for the delay.
- I do think this is at least AI-assisted, for editing and/or source summarization. I don't think it was fully AI-generated. This seems to apply mostly to the 2025-26 edits; obviously, the edits before 2023 don't count, but by definition everyone who uses AI now was not using AI in 2018, so that's irrelevant either way.
- Anyway, the clusterfuck of "highlighting the blah emphasizing the whatever" is exactly how AI text would come out around this time period. The source-text integrity isn't too bad compared to some other instances I've seen, the discrepancies I've found are more subtle:
- In general -- which is the case with AI text more often than not -- a lot of the stuff that is claimed to be "highlighted"/"emphasized" is not actually highlighted or emphasized, just mentioned. For instance,
highlighted the game's visual appeal as among the prettiest on the platform
is referring to a brief aside in a sentence about something else. Words have meanings. Virgin's "Digicel" process, a marketing term for a suite of tools and techniques for digitizing hand-drawn animation, aligned with Disney's vision for a game that felt like an extension of the film.
The "aligned with blah" is typical AI phrasing, and synthesis beyond what the source mentions: one interviewee speculates that Disney chose Virgin because John and Andy's team's work, but doesn't mention the vision, and it's also unclear whether it was marketed as "Digicel" at the time.
- In general -- which is the case with AI text more often than not -- a lot of the stuff that is claimed to be "highlighted"/"emphasized" is not actually highlighted or emphasized, just mentioned. For instance,
- Gnomingstuff (talk) 21:22, 7 March 2026 (UTC)
- I have done some LLM patrol/cleanup and a few GA reviews. I also suspect there was LLM involvement in the large content additions in February . The tone is both reminiscent of AI-assisted prose and different in style from CT's earlier edits to the article as well as clearly human-written contributions they made to other articles . I will do a deeper analysis tomorrow (including extensive source-to-text integrity checks) and will provide a third opinion once done. NicheSports (talk) 06:25, 8 March 2026 (UTC)
- Thank you for pointing out those edits for comparison @NicheSports. I was wondering if you've had an opportunity to analyze this any further? Also, as you've done a few GA reviews in the past, how would you recommend I proceed? I'm thinking I should do a closer inspection of sources that I spot-checked. As Gnomingstuff pointed out in at least one instance, a brief aside in a source was characterized as "highlighted", which isn't accurate. Chao Garden 🌱 (hi) 06:18, 11 March 2026 (UTC)
- I am about 2/3 of the way through my analysis, it has taken some time. I should be able to post it today NicheSports (talk) 13:04, 11 March 2026 (UTC)
- Thank you for pointing out those edits for comparison @NicheSports. I was wondering if you've had an opportunity to analyze this any further? Also, as you've done a few GA reviews in the past, how would you recommend I proceed? I'm thinking I should do a closer inspection of sources that I spot-checked. As Gnomingstuff pointed out in at least one instance, a brief aside in a source was characterized as "highlighted", which isn't accurate. Chao Garden 🌱 (hi) 06:18, 11 March 2026 (UTC)
- Sure, no problem. Gnomingstuff – Your knowledge in identifying likely AI-generated content has been requested, if you wouldn't mind taking a look at Disney's Aladdin (Sega Genesis video game) and giving your thoughts. Mr rnddude (talk) 16:59, 6 March 2026 (UTC)
- Potentially. There are passages in the prose that raise my eyebrows and I am also at least a little wary of large individual edits, but these are spaced apart by multiple hours indicating an editor actively working on it before posting. For example on 14 Feb 2026 we have edits at 01:51 (+2,855 bytes), 05:37 (+3,794 bytes) and, 08:53 (+5,567 bytes). It is not unreasonable to be able to build any of those edits over the course of a couple hours and that is an indicator of human-made content. I'm inclined to give benefit of the doubt. I can also ping an editor that I have crossed paths with that works heavily in dealing with AI generated content on Wikipedia if they're willing to give their considered opinion if that'd be helpful. Mr rnddude (talk) 10:49, 6 March 2026 (UTC)
- Disney's Aladdin (Sega Genesis video game) Chao Garden 🌱 (hi) 08:44, 6 March 2026 (UTC)
The above reads like a witch hunt. There is only a single WP guideline covering AI: WP:NEWLLM which consists of a single sentence: "Large language models (LLMs) can be useful tools, but they are not good at creating entirely new Wikipedia articles. Large language models should not be used to generate new Wikipedia articles from scratch" [emphasis added]. The only WP document mentioning disclosure of AI use is WP:LLM which is merely an essay. WP:AISIGNS is merely an advice page. GA reviews should not be relying on essays when evaluating GA criteria. If the article has factual inaccuracies or peacock phrases, the reviewer should: (1) Ask the nominator review all AI-like phrases and remedy them, if needed; and (2) Perform a thorough spot check on those phrases, by asking the nominator to supply quotes from the sources that support the phrases. The GA Talk page is not the place to promote WP:LLM from essay to policy. Noleander (talk) 14:55, 11 March 2026 (UTC)
- Hi Noleander, that is actually exactly the approach I am taking... focusing on source to text integrity analysis (GA criteria #2). I am also planning to post this analysis on the GA review page, not here. Hopefully that resolves some concerns. NicheSports (talk) 15:40, 11 March 2026 (UTC)
- Continuing this entire discussion in the GA Review talk age is a great idea. Noleander (talk) 15:49, 11 March 2026 (UTC)
- In case anyone from this thread wants to add their thoughts, I posted the analysis at the GA review talk page NicheSports (talk) 07:01, 13 March 2026 (UTC)
- Continuing this entire discussion in the GA Review talk age is a great idea. Noleander (talk) 15:49, 11 March 2026 (UTC)
The above reads like a witch hunt
is an accusation that is both spurious and incendiary. Nowhere is the nominator being attacked, and raising a concern about the possibility that AI generated errors or misrepresentations appear in a potential GA is both well within the bounds as a subject for this venue and certainly does not constitute witch-hunting. Mr rnddude (talk) 17:11, 11 March 2026 (UTC)
Sync number of reviews and GAs to new username
My total reviews and GAs are not synced to my new username from what I can see. My old username was "Surayeproject3." Is there anyway that this can be changed so that the numbers are synced? Thanks. PresentlySuraye3 (talk) 22:48, 10 March 2026 (UTC)
- @PresentlySuraye3 see the instructions at User:ChristieBot#What to do if your username changes JuniperChill (talk) 20:04, 11 March 2026 (UTC)
Improperly filed GA nomination
There's an improperly filed GA nomination at Talk:Joe Hathaway (American Politician)/GA1 that was done by the page's creator, who also made it so they're reviewing their own nomination. I think it constitutes as a quickfail but I'm not sure on how to do it when they've already started the review, could someone take a look at this? reppoptalk 03:54, 13 March 2026 (UTC)
- I closed it as a quickfail. Generalissima (talk) (it/she) 04:38, 13 March 2026 (UTC)
Closing procedure question
Is there something else that should have been done with this old, failed GA nomination? It appears on the V&A Talk page as if the review is still in progress (i.e., there is no closing box or different background colour to indicate that the matter is no longer current). I'm not well versed in GA review procedures, but it seems that there may be a parameter that's missing somewhere. Cheers, Cl3phact0 (talk) 08:35, 14 March 2026 (UTC)
- Nothing really needs to be done. This review was probably done well before all that stuff was created. You could archive it if you wanted, since it's old, but otherwise there's no need to be adding templates to 13 year old review pages. ♠PMC♠ (talk) 09:12, 14 March 2026 (UTC)
- There are many reviews still done today without the archive formatting. ~~ AirshipJungleman29 (talk) 09:43, 14 March 2026 (UTC)
Talk:Cairn (video game)/GA1
ChatGPT review again
Talk:ChatGPT/GA2 was started by a user who is new to the GA process. I also think some parts are LLM-generated and it's very short. Could it be vacated? HurricaneZetaC 02:03, 16 March 2026 (UTC)
- Can you clarify: When you say "it is short" are you referring to the article or the review? When you say "some parts are LLM generated": article or review? When you say "a user who is new to the GA process" do you mean the nominator or the reviewer? Noleander (talk) 02:37, 16 March 2026 (UTC)
- It is evident Zeta is talking about the reviewer. Tbhotch™ (CC BY-SA 4.0) 02:54, 16 March 2026 (UTC)
- Alright. Well, there is certainly something unusual happening with that article: the first GA review was a few weeks ago, and the first reviewer passed GA with a cursory review. Nominator has a GA award on their page. But the GA promotion is not shown in the article Talk page Milestones section? Now it is undergoing a 2nd GA review? This is above my pay grade. Noleander (talk) 03:01, 16 March 2026 (UTC)
- @Noleander that’s because the first review was vacated after WT:Good article nominations/Archive 37#ChatGPT review. I’m asking for the same thing to happen here, as I don’t think that this review was proper. @Pietrus1 already left a talk page message to the user explaining the issue. ~2026-16563-32 (talk) 03:14, 16 March 2026 (UTC)
- Alright. Well, there is certainly something unusual happening with that article: the first GA review was a few weeks ago, and the first reviewer passed GA with a cursory review. Nominator has a GA award on their page. But the GA promotion is not shown in the article Talk page Milestones section? Now it is undergoing a 2nd GA review? This is above my pay grade. Noleander (talk) 03:01, 16 March 2026 (UTC)
- It is evident Zeta is talking about the reviewer. Tbhotch™ (CC BY-SA 4.0) 02:54, 16 March 2026 (UTC)
- I agree that the review should be vacated here. This article deserves an experienced reviewer. Generalissima (talk) (it/she) 01:29, 17 March 2026 (UTC)