Editing Eurovision Wiki:WikiProject AI Cleanup/Noticeboard (section)

== [[S. Saraswathi]] and [[Safia/Safdar]]  ==
{{AIC status
| cleanup_status = requested
| tracking_subpage = 
}} <!--  Valid statuses: requested, ongoing, unnecessary, completed. Update if needed!  -->

Both [[S. Saraswathi]] and [[Safia/Safdar]] show uses of AI including failed citation links and scoring 100% on AI detectors. The editor is question swears up and down that they don't use AI while they write at a 10th grade English level on my talk page then at a graduate level on the articles. I figure this is a good starting place but if they continue to use AI to write or edit articles then I'll have to take it up for community sanctions. '''[[User:Dr_vulpes|<span style="background:#4B0082; color:white;">Dr vulpes</span>]]''' [[User talk:Dr_vulpes|(Talk)]] 22:10, 15 March 2026 (UTC)

:{{ping|AlphaCore}} courtesy ping the editor who approved these drafts. To me there is partial evidence of being AI generated (mostly in how the prose provides interpretations of cited sources rather than summarizing what they say). ~[[User:Anachronist|Anachronist]] (who / me) <small>([[User talk:Anachronist|talk]])</small> 03:01, 16 March 2026 (UTC)
::Yeah the editor in question is claiming that if articles go though AfC then they are proof there was no AI used. They also deleted all of the AI use warnings from their talk page. '''[[User:Dr_vulpes|<span style="background:#4B0082; color:white;">Dr vulpes</span>]]''' [[User talk:Dr_vulpes|(Talk)]] 03:03, 16 March 2026 (UTC)
:::I have noticed that many reviewers on Wikipedia aren't familiar with [[WP:AISIGNS]]. That's understandable, because that document is fairly new and constantly being updated. To me, however, these articles look borderline, likely human-written with AI "polishing". ~[[User:Anachronist|Anachronist]] (who / me) <small>([[User talk:Anachronist|talk]])</small> 03:08, 16 March 2026 (UTC)
::::What's funny is the editor who made those edits just 100% lied about the use of AI, over and over and over again. I thought it went the other way, AI first then human muckery, but the outcome is the same. If the editor had just said they used AI to clean it up that would have been that. If you want to see how this all played out see my talk page. '''[[User:Dr_vulpes|<span style="background:#4B0082; color:white;">Dr vulpes</span>]]''' [[User talk:Dr_vulpes|(Talk)]] 03:24, 16 March 2026 (UTC)
:::::I mean, there are serious issues either way. The first sentence I checked, in Critical Reception, was {{tq|Arpita Sarkar of OTTplay rated the film 3.5 out of 5 stars and described it as a socially relevant drama examining the challenges faced by women working in male-dominated professions}}. Almost none of that is [https://www.ottplay.com/news/safia-safdar-out-on-ott-where-to-watch-plot-details-and-more-about-naseeruddin-shah-kanwaljit-singhs-social-drama/117546d17e773 verified by the source]: this isn't a critic, the writer does not appear to have given it a star rating, does not mention male-dominated ''professions'' specifically, etc. Which, plus the (relatively mild but consistent with GPT-5) phrasing, suggests that AI was used, because otherwise where the hell do you get that? [[User:Gnomingstuff|Gnomingstuff]] ([[User talk:Gnomingstuff|talk]]) 15:46, 16 March 2026 (UTC)
::@[[User:Anachronist|Anachronist]] My bad, I supposed to clean the article S. Sarswathi during or after the acceptance. What I did was that added two critics from reliable sources as I found them. But, I missed the part to do clean up as well. I aware about the LLM contents, however, I take this one in my mind and will be checking more carefully on AI. Thanks for pinging me. 
::And did I accept Safia Safdar? I believe I did draftify as it was lacking RS. [[User:AlphaCore| <b><span style="color:#0066FF">A<span style="color:#0099FF">l<span style="color:#00CCFF">p<span style="color:#00EEFF">ha</span>C</span>o</span>re</span></b>]] [[User talk:AlphaCore|talk]] 23:29, 16 March 2026 (UTC)
:::You didn't accept it. Nobody did. What apparently happened is that both the mainspace and draft versions coexisted, and {{U|Izno}} merged them together in mainspace.[https://en.wikipedia.org/w/index.php?title=Special:Log&page=Draft%3ASafia%2FSafdar] In my view the merge should have been done in the opposite direction. This should be re-draftified. ~[[User:Anachronist|Anachronist]] (who / me) <small>([[User talk:Anachronist|talk]])</small> 02:49, 17 March 2026 (UTC)
::::[[Special:MergeHistory]] works in only one direction: to the page that came later. I leave moves to and fro to the people who think they know the right final place. [[User:Izno|Izno]] ([[User talk:Izno|talk]]) 02:54, 17 March 2026 (UTC)
:::::Ugh. I forgot about that feature; I don't use it because it rarely works properly for me. I just move all the revisions to a new location, and then undelete the revisions that got overwritten.
:::::In any case, I have re-draftified the article. ~[[User:Anachronist|Anachronist]] (who / me) <small>([[User talk:Anachronist|talk]])</small> 02:57, 17 March 2026 (UTC)
::::::Using MergeHistory makes my (and any admin's) intention clearer. And coincidentally ensures that any choices to override [[WP:PARALLEL]] are deliberate. [[User:Izno|Izno]] ([[User talk:Izno|talk]]) 16:42, 17 March 2026 (UTC)
::::::: (offtopic) I have a patch in review to ensure that overriding [[WP:PARALLEL]] by undeleting has to be done deliberately too: [[gerrit:1248956]]. [[User:Pppery|* Pppery *]] [[User talk:Pppery|<sub style="color:#800000">it has begun...</sub>]] 00:21, 18 March 2026 (UTC)
:::::::Typically when I merge articles by moving one over another, there is usually no issue with parallel edits because there is no overlap in dates. Or if there is, the overlap is only a couple of edits. If I see a lot of overlap, I manually select the deleted edits to restore.
:::::::The explanation that MergeHistory works only one way would explain why I have never been successful in using it to draftify articles where a draft already exists. ~[[User:Anachronist|Anachronist]] (who / me) <small>([[User talk:Anachronist|talk]])</small> 04:13, 18 March 2026 (UTC)