Jump to content

Wikipedia:Wikipedia Signpost/2026-02-17/In the media

Add links
From Wikipedia, the free encyclopedia
The Signpost
← Back to Contents
View Latest Issue


File:PSM V10 D562 The hindoo earth.jpg
Unknown (The Popular Science Monthly 1877)
Public domain
300
In the media

Global powers see Wikipedia as fundamental target for manipulation

Contribute   —  
Share this
By Bri, Andreas Kolbe, HaeB, and Bluerasberry

The Times says Wikipedia was "hacked" and calls it an "important victory" for Jeffrey Epstein

TKTK
"Hacking" Wikipedia: Al Seckel in 2009

The Times (UK) reports "How Epstein scrubbed 2008 sex crime conviction from the internet" by "hiring a team of hackers to remove negative information about him on Wikipedia and Google". It quotes Epstein's PR team telling him:

Wikipedia was an important victory, as it will always be at the top of the search engine results. [...] We have stopped the hacking on your wiki site, and that was a major effort. Your wiki entry now is pretty tame, and bad stuff has been muted, bowlerized [sic], and pused [sic] to the bottom.

The edits described by The Times (examples: link, link, link, link; removal of the sex offender category was reverted by JodyB) were performed by an IP, 71.165.127.242. According to The Times, the work was done by Al Seckel, the husband of Ghislaine Maxwell's sister Isabel.

The Times describes Seckel as "a writer and self-styled expert in optical illusions who co-founded a group called the Southern California Skeptics that investigated science's relationship to the paranormal" and who was found dead "in mysterious circumstances" in 2015.

This "hacking" of Epstein's Wikipedia article appears to have been quite remunerative:

Seckel chased up payment for his work and Epstein complained about the cost. "I was never told never, that there was a 10k fee per month„ you inittaly [sic] said the project would take 20.. then another 10. then another 10" he wrote in one message.

Note that The Times is a little late to the party here – other outlets covered the same story a few months ago, and readers can find a more comprehensive summary of the edit history of Epstein's article in the Disinformation report published in The Signpost's 1 December 2025 issue.

It is not immediately clear why The Times chose to describe people making (later-undone) edits to a publicly editable website as "hacking", seemingly on the basis that said people described themselves as such in their own marketing materials.

For Wikipedia references in the latest set of Epstein files released on 30 January 2026 see this issue's Disinformation report. – AK, J

Hand-marked copy of Trump's Wikipedia article in Epstein files

In an article titled "Epstein jail cell pics and Trump Wikipedia page included in newly released files", The Independent provides a link to the pdf of a hand-marked printout of Trump's Wikipedia article, dated 22 May 2022.

The printout contains many manually underlined passages and a small number of hand-written comments. Some words in the Wikipedia article itself – among them "d'Italia", "Stone's", "D'Souza", "d'affaires", "Lesley", abbreviations like "p.m.", "U.N." and "D.C.", and parts of its reference URLs – are redacted in the pdf and hidden behind black bars. – AK

Chinese Wikipedia competitor

"BaiduWiki" went live around 9 February, with English and other editions AI-translated from Baidu Baike, the South China Morning Post reports in "China's Baidu unveils AI-driven Wikipedia challenger in bid for international users".

Current usefulness of the site seems rudimentary to non-existent; an English-language search for "Donald Trump", for example, redirects to BaiduWiki's entry Miss Universe Organization, while a search for "Goethe" takes users to BaiduWiki's entry Translated Literary Classics: Faust, on "a book published by Shanghai Translation Publishing House in 2013". – AK

Not a fan of UC Berkeley project

Campus Reform claims "A University of California, Berkeley professor tasks students with editing and creating Wikipedia articles about 'queer and trans people of color' instead of taking final exams" and has screenshots of WikiEd dashboards included in their piece. The last issue of The Signpost noted reporting from The Daily Californian [1] of the coursework in "300,000 edits, 3,000 refs, 96 million views", but more recent reporting also comes from SFist [2], Them [3], The College Fix [4], UC Berkeley News [5], and Queer News Tonight [6]. If you're wondering what the problem is, let's just say Campus Reform took issue with some of the topics the students chose, and maybe the fact that the professor "frames the project as a form of opposition to the Trump administration". – B, BR

Wikipedia, but with more X appeal

Xikipedia is a web-based Simple English Wikipedia reader that presents an environment that many media noticed and called "doomscrolling" or other related terms. Coverage included AV Club, Boing Boing, Engadget, Gizmodo, Ground News, Hacker News, Let's Data Science, and Stuff (South Africa).

We weren't sure what to make of it at The Signpost until the author, Wikipedian Rebane2001, contacted the Signpost team pre-publication. They said the project is "art/commentary on modern social media algorithms", that it's different from the infinite scrolling, but random WikiTok by virtue of having a "feed [that] adjusts based on which 'posts' you like", and pointed us to this GitHub readme for more details. – B

We are (self) aware that it's lists, all the way down

TKTK
You may find this image in an article encompassed by List of English-language metaphors.

Boing Boing found list of lists of lists – containing entries such as lists of Atlantic hurricanes, lists of Category 5 hurricanes, lists of tornadoes and tornado outbreaks, lists of physics equations, lists of metalloids, lists of celebrities, lists of centenarians, lists of deaths, lists of ethnic groups, and lists of LGBTQ people – is "technically practical but also function[s] as [a] conceptual joke" showing how Wikipedia is "obsessively organized, slightly absurd, and self-aware enough to know it."

In a Russell's paradox fashion, the list of lists of lists contains itself, a fact noted by Boing Boing. – B

Autobiography of a WMF advert

WP25, "Knowledge is Human" (text CC-BY-SA 4.0, attributions here)

In Communication Arts, the ad agency kin explains how the "Knowledge is Human" campaign, which they call a docuseries, was developed for Wikimedia Foundation.

See prior Signpost coverage of the 25th anniversary, including some bits about the Foundation's own promotion. – B

In brief

Politics
Policy
  • Simple Summaries controversy revisited: "At Age 25, Wikipedia Refuses to Evolve – The digital commons champion faces a crisis of its own making", says Wikipedian and academic Dariusz Jemielniak in IEEE Spectrum.
  • Section 230 turns 30: The Electronic Frontier Foundation published an interview with Jacob Rogers from the Wikimedia Foundation's legal team about the importance of Section 230 for Wikipedia, titled "The Internet Still Works: Wikipedia Defends Its Editors". In response to the question "If you had to describe to a Wikipedia user what Section 230 does, how would you explain it to them?", Rogers stated: "If there was nothing—no legal protection at all—I think we would not be able to run the website. There would be too many legal claims, and the potential damages of those claims could bankrupt the company. Section 230 protects the Wikimedia Foundation, and it allows us to defer to community editorial processes. We can let the user community make those editorial decisions, and figure things out as a group—like how to write biographies of living persons, and what sources are reliable. Wikipedia wouldn’t work if it had centralized decision making." The interview appears to have been published on occasion of the law's 30th anniversary earlier this month, but also comes at a time where activists have raised alarm about serious efforts going on to repeal or "sunset" it, via bipartisan legislation in the US Congress.
Technology
Is AI's relationship to the infosphere like that of a blood sucking tick to mammals? We're about to find out what happens when creators square off against "an extraction layer that absorbs creative work without preserving its lineage or sharing in the value it creates".
  • Don't worry about that weird feeling, it's just data extraction: "where creativity may be heading next" needs to be shaped by an upside for trust-based platforms like Wikipedia and revenue-sharing models like YouTube's – but faces problems with "an [AI] extraction layer that absorbs creative work without preserving its lineage or sharing in the value it creates". (Muse by Clios, "What AI Can Learn from YouTube and Wikipedia").
  • Slopping Stopping the slop-generator: Wikipedians building works in languages other than English are "both populating and fighting the world's regional language AI engines" (Rest of World, "The volunteer Wikipedia army protecting against AI slop").
  • Love it or hate it?: A debate over the continued suitability of archive.today was reviewed by Ars Technica. The focus of the story is Wikipedians' reactions to a claim that archive.today's operator used the site to launch a distributed denial of service (DDOS) attack from its users' browsers.
  • Why do you assume you're the smartest in the room, Grok?: Whereas Columbia Journalism Review says Wikipedia's text comes from "an army of humans [who] provide clarification and updates, in dialogue with one another", a media researcher at the University of Massachusetts Amherst calls challenger Grokipedia "a vertically integrated, centrally controlled knowledge production system" and its "decision-making dialogue [is] with itself".
  • Human-Centered AI and Wikimedia Enterprise: A Wikimedia Enterprise executive spoke at a Stanford University Human-Centered AI (HAI)/data science conference. (The Stanford Daily, "Wikimedia explains combating AI in HAI seminar")
Society
Fun



Do you want to contribute to "In the media" by writing a story or even just an "in brief" item? Edit next week's edition in the Newsroom or leave a tip on the suggestions page.


← Previous "In the media"
+ Add a comment

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

Delighted to see List of lists of lists getting another mention! It is certainly true that both great powers and "educational" editing can cause significant problems. I think we are somewhat alert to the former, but less so to the latter. ~~

You could wikilink archive.today. And Ars Technica. Gråbergs Gråa Sång (talk) 09:43, 17 February 2026 (UTC)[reply]

  • It is not immediately clear why The Times chose to describe people making (later-undone) edits to a publicly editable website as "hacking"

Because that's how it was described in the e-mail with Epstein, regardless of whether it wasn't actual hacking. You put the [citation needed] note as a bad joke, but the media is focused on what Epstein stated. – The Grid (talk) 14:00, 17 February 2026 (UTC)[reply]

Also, "hacking" is sometimes used in a less than formal sense. Gråbergs Gråa Sång (talk) 16:11, 17 February 2026 (UTC)[reply]
  • Xikipedia with category:science as selected interest
    Xikipedia is a web-based Simple English Wikipedia reader … We weren't sure what to make of it Love that webapp! Re "what to make of it" – you don't need to and maybe shouldn't add your subjective opinions about such projects and could just report on it. I think this format is great and allows casually scrolling through some articles to discover articles about diverse topics while on a train for example. Obviously, I don't want to see boring articles I find fully irrelevant to the world and of no interest to me so this is much more useful than a feed showing random articles for example. Since it only uses Simple EN WP and is not available as a native app which I prefer using, I'm currently using the Wikipedia Android app's recommended reading list feature instead which however has just 20 articles at max to go through based on interests that I were allowed to select. If you like the Xikipedia concept / prototype, check out proposal W506: Discovery feed in Wikipedia app similar to social media apps (incl article segments, vids, …) (voting open). I think it could substantially increase Wikipedia reads and let more people waste less of their time and instead use that time to learn new things.
    Audio version of List of lists of lists
    --Prototyperspective (talk) 19:35, 17 February 2026 (UTC)[reply]
  • This is the first Signpost article I read. Are they always this good? The authors have a way of mixing serious, engaging topics (I had no idea that authors were paid out to obscure information about Epstein!) and cold, witty sarcasm. It's entertaining and topical, while still opinionated enough to open up discussion. Going to read these every time now. Super cool. Octostomp (talk) 01:56, 18 February 2026 (UTC)[reply]
    Yes, always. Just kidding, we try to make it engaging and readable yet on topic for our readers (there's a whole team here, not just those of us on the byline). Sometimes it comes out better than others. Glad it worked for you. ☆ Bri (talk) 06:51, 18 February 2026 (UTC)[reply]
  • I'm surprised and disappointed at the lack of context in that mention of the Algemeiner article:

    The author is a contributor to HonestReporting, a Jerusalem-based media watchdog with a focus on antisemitism and anti-Israel bias.

    Hardly objective reporting about "Wikipedia's Israel problem"… — OwenBlacker (he/him; Talk) 08:17, 18 February 2026 (UTC)[reply]
    The author's affiliation is mentioned above and the site wikilinked. I thought that was enough. Andreas JN466 23:49, 22 February 2026 (UTC)[reply]

Commenting on the Spectrum take on Simple Summaries -- beyond its blatant disregard for any of the community's wishes, it treats younger people like mindless consumers of whatever new tech is pumped out. While it may be true that Gen Z reads more short-form content, Wikipedia articles at their best can provide such kernels of information in their structure (such as with infoboxes and lede sections) and more importantly, Gen Z hates AI. ✨ΩmegaMantis✨(he/him) ❦blather | ☞spy on me 16:45, 18 February 2026 (UTC)[reply]

@OmegaMantis IMO if anything is going to attempt to bring down Wikipedia, it's large governments and corporations trying to withhold information from the public. Not whatever nonsense they said. Octostomp (talk) 17:16, 18 February 2026 (UTC)[reply]
I don't like LLMs overall and am not supporting those simple summaries but if Wikipedia "can" does not mean Wikipedia 'does'.
Lots of articles have lede sections that are too complicated and/or long. Lots of people, especially GenZ but not only those, are for example using AI summaries and LLM chats instead of Wikipedia because they're used to too-often too complicated ledes. These are millions of lost reads/readers who now instead from Wikipedia get their info from accuracy-agnostic machines built by "corporations" or, since this has been a problem before genAI already, low-quality websites. Prototyperspective (talk) 19:07, 18 February 2026 (UTC)[reply]
Definitely true. I think along with simply making shorter lede sections, we also need to advertise and better curate our sister projects. There's probably a large amount of people who only wanted a definition but got a whole article that would rather check Wiktionary, for example. Poor Wikinews gets more views when linked to Wikipedia articles, and I think it would definitely help people to have easier ways to access other Wikimedia content on Wikipedia, if not help the communities of our projects next time more people are hungry for a shut-down. ✨ΩmegaMantis✨(he/him) ❦blather | ☞spy on me 20:30, 18 February 2026 (UTC)[reply]
It's your Signpost. You can help us.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Wikipedia_Signpost/2026-02-17/In_the_media&oldid=1340476618"