The Signpost

File:L Sanger.jpg
Larry Sanger
CC BY-SA 2.5
0
34
300
News and notes

Larry Sanger returns with "Nine Theses on Wikipedia"; WMF publishes transparency report

Contribute   —  
Share this
By HaeB and Bri

WMF publishes transparency report for January–June 2025

TKTK
"Origination of user information requests" by country (from the report)

The Wikimedia Foundation has published its transparency report for the first half of 2025, about "requests we receive to alter or remove content from the projects, and to provide nonpublic information about users."

The section on requests for user information ("such as IP addresses or user agent information") reports that 64 user accounts were "potentially affected" by such requests, but only one "actually affected". Among the 20 government requests received, the largest number (8) came out of India. The Foundation "partially complied" with only one request, out of France, and "fully complied" with 0. This was down from 2 granted requests in the second half of 2024: one from Brazil ("fully complied") and one from India ("partially complied"). The latter had received a great deal of community attention, including an open letter with the largest number of signatures in Wikimedia history (Signpost coverage).

The Foundation proudly points out that "Compared to other companies, we received relatively few requests, and granted relatively low percentages", citing numbers from LinkedIn, Meta and X (formerly Twitter), who during a comparable recent half-year timespan granted 723, 251,028 and 10,581 requests for user information, respectively.

A look at the previous transparency reports from the last half decade (the report for the first half of 2019 seems to have been removed or never published) confirms that such low numbers are the norm for the Wikimedia Foundation - although the second half of 2023 seems to have been an outlier, in that no less than 896 user accounts were "actually affected" by the 5 requests granted:

timespan Total requests Requests granted User accounts potentially affected User accounts actually affected
July to December 2019 35 2 ? ?
January to June 2020 30 1 72 2
July to December 2020 32 2 3,119 4
January to June 2021 30 3 38 5
July to December 2021 18 1 33 2
January to June 2022 31 0 38 0
July to December 2022 29 0 3,816 0
January to June 2023 41 3 8,712 4
July to December 2023 32 5 985 896
January to June 2024 26 2 186 2
July to December 2024 23 2 33 4
January to June 2025 30 1 64 1

The Foundation's transparency report also provides other kinds of information, e.g. about "Requests for content alteration and takedown", or about "Orders from EU Member States" that it received under the EU's Digital Services Act (DSA).

H

Larry Sanger is "baaaaack!" with "Nine Theses on Wikipedia"

Larry Sanger has been largely inactive as a Wikipedia editor since his departure in 2002 as Wikipedia's "chief organizer" who (as employee of Jimmy Wales) had crafted several of its core policies. On September 29, he updated his user page to announce that

I'm baaaaack! [...] For most of 2025, I have developed Nine Theses on Wikipedia, which is partly an extended criticism and partly a reform proposal. Unlike much of my previous writing and speaking about Wikipedia over the past 20 years or so [cf. Signpost coverage], this is not merely negative. It is a realistic plan to make Wikipedia better. I hope you will take it seriously.

The nine theses, expanded upon in great detail in the document (which "is 37,000 words, something like a 150 page book" according to Sanger), are:

"1. End decision-making by 'consensus.'"
Sanger argues that Wikipedia's "notion of 'consensus' [cf. WP:CONSENSUS] is an institutional fiction, supported because it hides legitimate dissent under a false veneer of unanimity."
"2. Enable competing articles."
Sanger proposes that "Wikipedia should permit multiple, competing articles written within explicitly declared frameworks, each aiming at neutrality within its own framework", because "Wikipedia is now led by [...] uncompromising editors. As a result, a favored perspective has emerged: the narrow perspective of the Western ruling class, one that is 'globalist,' academic, secular, and progressive (GASP). In fact, Wikipedia admits to a systemic bias, and other common views are marginalized, misrepresented, or excluded entirely."
"3. Abolish source blacklists."
A criticism of Wikipedia's "Perennial sources" page (which serves to summarize community consensus about the reliability of frequently discussed sources). In particular, Sanger objects to its treatment of some specific news publications on the US political right: "Wholly 'deprecated' sources include, for example, Breitbart, the Daily Caller, and Epoch Times. 'Generally unreliable' outlets include much of Fox News reporting and all of the New York Post and The Federalist [...]".
"4. Revive the original neutrality policy."
Sanger argues that "The present policy on neutrality [WP:NPOV] should be revised to clarify that articles may not take sides on contentious political, religious, and other divisive topics, even if one side is dominant in academia or mainstream media. Whole parties, faiths, and other 'alternative' points of view must no longer be cast aside and declared incorrect, in favor of hegemonic Establishment views."
(Earlier this year, a two-episode podcast interview of Sanger with the Discovery Institute had highlighted the Wikipedia article on intelligent design as an alleged example of such failings.)
"5. Repeal 'Ignore all rules.'"
Sanger relates how he had posited this "humorous rule" himself back in 2001 "to encourage newcomers. Ironically, my joke now serves to shield insiders from accountability" in its present form (WP:IAR).
"6. Reveal who Wikipedia’s leaders are."
Sanger holds that "the Wikipedia users with the most authority)—'CheckUsers,' 'Bureaucrats,' and Arbitration Committee members [...] *should* be identified by their real and full names, so they can be held accountable in the real world."
(On the talk page, he clarified that he does not "support doxxing people who rely on their anonymity in the system", decrying as inaccurate a media report from earlier this year which had implied that he was supporting such efforts.)
"7. Let the public rate articles."
"8. End indefinite blocking."
("Indefinite blocks should be extremely rare and require the agreement of three or more Administrators, with guaranteed periodic review available.")
"9. Adopt a legislative process."
Sanger argues that this is needed because "Wikipedia’s processes for adopting new policies, procedures, and projects are surprisingly weak. [...] Incremental policy tweaks cannot deliver the bold reforms Wikipedia needs. No clear precedents exist for adopting significant innovations. The project is governed by an unfair and anonymous oligarchy that likes things just as they are."
Somewhat surprisingly, this is also the only part in the entire document where Sanger - very briefly - mentions Citizendium, the wiki-based online encyclopedia he launched in 2006 (initially as a fork of Wikipedia), and which intentionally deviated from Wikipedia in several ways that seem consistent with his current theses - such as a real name policy for all contributors, or a "community charter" with "legislative authority" (Signpost coverage: "Citizendium adopts charter, Larry Sanger's leading role ends"). As this Signpost writer argued in a talk at Wikimania 2009 ("Lessons from Citizendium"), the project can thus be seen as a "long-time experiment testing several fundamental policy changes, in a framework which is still similar enough to that of Wikipedia to generate valuable evidence as to what their effect might be on [Wikipedia]". But in the lengthy rationales for his nine theses, Sanger unfortunately fails to cite any learnings from his several years of efforts to make Citizendium succeed as its editor-in-chief - or from the various other encyclopedic projects he has worked on since his departure from Wikipedia.

Various Wikipedians have so far commented on the talk page and in a village pump thread.

Sanger also announced his theses in an article at The Free Press (see "In the media" in this issue) and in a thread on Twitter/X, where he added:

Wikipedia could change. It's not impossible.

But only if you make a lot of noise both on social media and on Wikipedia itself. The current narrative is controlled by a few hundred people. What if 1000s (politely) descended on Wikipedia?
These are, in fact, very reasonable, commonsense proposals from anybody's point of view. We can put pressure on Wikipedia at all levels to adopt them. If they do nothing or refuse to change, there will be consequences.

I am Tucker Carlson's interviewee today—we talked about both criticisms and this reform proposal.

The 93 minute interview with Tucker Carlson touched on topics such as Carlson's theory about "Wikipedia’s Dark Alliance With Google" - alleging the existence of "a deal with Google that allows them to be the top search result" (Sanger agreed that "you very well could be right", but offered the alternative theory that Wikipedia might in its early days have benefited from a "feedback loop" with Google's algorithm, by being the first website to cover various topics).

Other parts of the interview caused "MAGA [to] Melt[] Down Over Wikipedia ‘Blacklist’", as summarized by The Daily Beast. These reactions included Elon Musk announcing that at his company xAI, "We are building Grokipedia [..] Will be a massive improvement over Wikipedia". Sanger reacted wearily: "Let’s hope it won’t be as biased as Grok itself."(Several weeks earlier, Musk had commented on the All-In podcast about possibly using Grok to "rewrite Wikipedia to remove falsehoods and add missing context". See also earlier Signpost coverage of Musk's grievances: "Op-ed: Elon Musk and the right on Wikipedia", "Wikipedia is an extension of legacy media propaganda, says Elon Musk")

In a 2013 tweet, Sanger had announced that "I am finished with Wikipedia criticism. Quote this back to me if I happen to lapse." Reform, however, was not mentioned. – H

Temporary accounts rollout soon

Editing with Temporary Accounts instructional video

Temporary accounts, formerly known as "IP masking", has been tentatively scheduled for rollout on English Wikipedia on October 7 (see announcement). The feature is already active on several Wikipedias and involves removing IP visibility for people who choose to edit without logging in.

For further reading on the discussions and rationale leading up to this, see previous Signpost coverage from 2020, 2024, and 2025. – B, H

Brief notes

This is a page from the Human Rights Impact Assessment's executive summary, featuring a "Recommendation on the Use of GenAI in Wikimedia Context"
Signpost
In this issue
+ 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.
I think this is unlikely. jp×g🗯️ 07:47, 2 October 2025 (UTC)[reply]

So, do we switch from "Hello IP!" to "Hello TA!"? Gråbergs Gråa Sång (talk) 11:37, 2 October 2025 (UTC)[reply]

Per WP:VPWMF#Update: two weeks more to prepare for the change, we've decided to roll temp accounts out on October 21st and communicate more about the change. @Bri, @HaeB, thanks for mentioning it in this issue! I will gladly help you prepare a more detailed note in the next one, to make sure that people not only aren't surprised, but also know how to deal with temp accounts efficiently. SGrabarczuk (WMF) (talk) 21:33, 6 October 2025 (UTC)[reply]

At this point, Sanger is reasonably considered an outsider and thus is bringing an outsider's views, which on a basic level I welcome, and we should see more of such. That said, following is my first blush take of Sanger's ideas:

  1. 1, 2, 3, 6 and 8 are totally out of the question for me. In my estimation as a longtime editor, they are just bad ideas. And I suspect Elon Musk and his ilk froth at the mouth over 6. No, Elon, you are not getting that information. No, no, no. I have always supported transparency in government, but this is not government, and there is no "government force" protection for our officers. On the other hand, I would encourage voluntary transparency, and would consider that courageous.
  2. On 4, perhaps Wikipedia could do a better job on presentation of other perspectives, but I also think an encyclopedia should reflect the prevailing view on what is factual or is predominantly held to be true at the time it is "printed" (i.e. media is a product of its time). Extensive coverage of minority positions due to them maybe eventually having more weight is prospective. And we don't do prospective writing. We write based on current knowledge and prevailing views. That is what an encyclopedia is. That all said, does Wikipedia have a Western bias? If you mean English Wikipedia, yes, and that's because English is a language that dominates the West. Good luck getting English writers to magically think and write outside Western perspectives. If you really want to talk about bias, though, the true bias of all of Wikipedia is corporate bias. As corporates swallow up most of the media we depend on for sourcing, we are unfortunately going to tend to reflect their POV, which is anti-labor and anything else not approved of by corporate owners. As a movement, we need ways to counterbalance this.
  3. On 5, I am open-minded to a community-wide review of IAR. I've come to appreciate that we have policies and guidelines and the expectation they be followed, as this helps lead to a cohesive, coherent overall product. Not following them on purpose should be considered extraordinary. Abrogating a policy or guideline outside of community consensus for that abrogation should be frowned upon. At the same time, we should have clear avenues and directions for challenging policies and guidelines... especially guidelines. So, if an editor defiantly wishes to write against a guideline, we can tell them: "Follow the guideline, or challenge the guideline – those are your only two choices. You may not ignore them."
  4. 7 actually sounds like a cool idea, although I don't think it should be a bare numerical rating but rather a set of ratings based on various factors we consider important, with the examples of "Well-sourced?" or "Balanced presentation?". But definitely no written feedback should be allowed as part of this – that would be a duplicative mess and a moderation nightmare. People should as always be directed to an article's talk page when they wish to have a discussion about an article's quality and/or how it may be improved.
  5. 9 has a kind of reason to it, but this isn't a democracy, it is a meritocracy. Weighting things down with a legislative process would work against GTD (getting things done). And due to the ongoing shortage of active editors, we need Wikipedians being WP:HERE, rather than "there", legislating.
  6. As for Sanger pushing 1000s to descend on Wikipedia politely, I'm guessing the audience actually enthusiastic about his overall ideas would be anything but polite and will tend to cause more damage than find ways to push for changes via consensus. The few who break through constructively are unlikely to change any consensus, of course, unless and until they are prepared to become active editors and eventually get to a place of understanding of what they are talking about on any given topic of relevance. It's like expecting hordes of the unwashed to storm libraries with an expectation of getting them to change away from their chosen classification method or alter material use policies. It's a weird call. If not unhinged.
  7. Re: "If they do nothing or refuse to change, there will be consequences." I hereby LOL at such false bravado, and retort "You and what army?" Change here is slow in many areas, but there is no refusal to change. Also, this is effectively a library catering to readership. The idea we need to have continual, rip-roaring change is rather antithetical to maintaining a calm atmosphere of spreading knowledge and evolving ways to enhance that. We are a body of thinkers and teachers. This is not sport. And we shall not kowtow to Goths, Vandals and X hooligans. Come here to work constructively, or we will (politely) show you the door.

Stefen 𝕋ower's got the power!!1! GabGruntwerk 19:49, 2 October 2025 (UTC)[reply]

Further, re: Grokipedia, I welcome the competition. But given the state of AI at this point, I don't see it as actual competition in the near term. No matter the compendium, if readers get a sense of not being given a factual presentation of the subject, that will cause challenges for that compendium. Wikipedia is obviously not immune from that, but unlike AI, we are or can be self-corrective. AI algorithms can be improved, but you have no human editors of a given subject discovering a problem (perhaps one AI can't detect) and then fixing it. And when Musk says "remove falsehoods and add missing context", I frankly snicker as someone who has watched Musk over the years. He is obviously wanting to apply a right-wing tilt and give a lift to pseudoscience or other widely rejected positions, just as he has done on X. He as the richest person ever wants everything to bend to his POV, and we are not doing that, and that makes him fume. Fume on, Elon. Stefen 𝕋ower's got the power!!1! GabGruntwerk 20:26, 2 October 2025 (UTC)[reply]
After reading a lot of the responses, I've decided to slightly soften my response on Sanger's #3, which I said was out of the question. I still think getting rid of the blacklist altogether is out of the question, but I do think it makes sense to have a periodic review of blacklist entries because 1) it's always possible a media outlet has made improvements to their journalistic efforts; 2) it's always possible the initial judgment to place a media outlet on the list was not well-founded and possibly didn't have a well-rounded group of discussion participants. The bottom line is we should be as fair as we possibly can be. A placement on the list should not be irreversible. Stefen 𝕋ower's got the power!!1! GabGruntwerk 02:04, 4 October 2025 (UTC)[reply]
One has to understand the essay within the context in 2025. It's not just an essay written by someone who wants to provoke discussion within the existing structures. It's something written by someone with a very specific agenda they've been very clear about for many years and which they've recently explained clearly on popular media. If we don't listen to that, then more fool us. JMWt (talk) 11:17, 3 October 2025 (UTC)[reply]
I basically agree. At the same time, I am not the kind of person who rejects ideas due to judging the motivation of the person bringing forth the ideas. Sometimes a person with impure motivation can produce ideas worth chewing on. Even if Sanger's goal was to ruin the Wikipedia, it's still possible he has tripped on ideas that actually would improve it. Stefen 𝕋ower's got the power!!1! GabGruntwerk 20:49, 3 October 2025 (UTC)[reply]
"That all said, does Wikipedia have a Western bias? If you mean English Wikipedia, yes, and that's because English is a language that dominates the West. Good luck getting English writers to magically think and write outside Western perspectives." I disagree, systemic bias (US/Europe [mostly the former], men, nerd culture and so on) is an entrenched issue that needs to be countered, the reason we have projects such as Women in Red in the first place. We of course cannot force voluntary editors to cover other topics but as a project can clearly encourage editors to do so in even more ways than we do now. Gotitbro (talk) 11:24, 3 October 2025 (UTC)[reply]
I mean, we are aware of our own bias, WP:BIAS (and the average Wikipedian). There's going to be a bias with any publication. But Sanger's arguments are a red herring. I see he loves to bring up WP:AGF for any comments about himself but when your life involves grifting, there's been no more rope to believe there's good intentions for a very long time. – The Grid (talk) 13:00, 3 October 2025 (UTC)[reply]
Of course we should have editors with as many perspectives as possible editing here. When you say "systemic bias is an entrenched issue that needs to be countered", we are in agreement, not disagreement. I was speaking to a specific bias - one that will be extraordinarily difficult to overcome because English is the predominant language of the West. Stefen 𝕋ower's got the power!!1! GabGruntwerk 21:03, 3 October 2025 (UTC)[reply]
1. "End decision-making by 'consensus.'"
The alternative would be, I guess, letting named arbitrators decide what goes in/out away from the community process [point 6]. How does this solve bureaucratic Nupedia problems for which Sanger and Jimbo created Wikipedia I am unsure.
2. "Enable competing articles."
We already have such articles, to take a leaf from the presented gripes:- Evolution and the Catholic Church, Islamic views on evolution, Jewish views on evolution, Hindu views on evolution etc. The question is how much weightage do you give to these on the main article of evolution and as an encyclopedia we simply cannot present these on equal footing with the scientific literature if the main article is under the topic of science.
3. "Revive the original neutrality policy. - Sanger argues that "The present policy on neutrality [WP:NPOV] should be revised to clarify that articles may not take sides on contentious political, religious, and other divisive topics, even if one side is dominant in academia or mainstream media."
From the interview, NPOV is the crux of Sanger's problems with enwiki. But beyond some real concerns of MOS:LABEL vios [and needless focus on controversies i.e. undue weight] at BLPs etc. in contravention of extant guidelines no less], I don't see problems with our policies as such. Editorial conduct and violations of P&G as such need to be resolved by participating in the editorial process. Even Ashley Rindsberg, whom Sanger references in the interview, realizes this when he critizes editors and the gaming of policies by them rather than the policies themselves.
4. "Abolish source blacklists."
Some RSN decisions/labels are questionable but that doesn't impeach the validity of RSP as a whole. [PS: Tucker's reaction to the list is the most funny part of the interview.]
5. "Repeal 'Ignore all rules.'"
Agree, while rookie edits and mistakes can be dismissed under such a notion. The usage of this is mostly on an ad-hoc basis for whatever one wants to do. We need to apply P&G without exceptions, standards need to be had.
6. "Reveal who Wikipedia’s leaders are."
Agree. For admins this can be encouraged but for ArbCom and other formal positions of authority this can be made policy.
7. "Let the public rate articles."
Agree, the Hebrew Wikipedia for instance has a public feedback form under ever article. [Also see my comment for 9.]
8. "End indefinite blocking."
Absolutely not. Certainly Sanger knows the trouble with interested government actors, PR firms and the like [as clearly stated by him multiple times in the interview], to think :that we can handle such a deluge without perma bans is fantasy.
9. "Adopt a legislative process."
We definitely can have a more participative process when advertising policy changes, banner messages simply don't cut it. Talk pages are unknown to the public and so on and so forth. This is an outreach failure (and for all the fearmongering that Ashley Rindsberg does, at least it puts things into perspective that normal people edit enwiki rather than a 'cabal' which controls it). But for all the more 'radical' changes sought by Sanger, he will have to participate in the extant processes to gain community inertia to change them, but saying that only 'named' accounts can weigh in on policy runs into the same problems as 1. Gotitbro (talk) 16:13, 3 October 2025 (UTC)[reply]
I agree that some legitimate alternate sources are blacklisted that I think should not be. Since I doubt that they all are legitimate, I think the list probably needs to be more limited, but not abolished. Some are blacklisted simply because those pursuing a point of view have convinced enough others of like mind that they should be.
Unless I am mistaken, neutral point of view allows alternate or minority views if properly sourced, may even encourage them, provided they are identified as such.
Ignore all rules should be removed. There may be exceptions to some rules or differing yet proper interpretations. Ignoring all rules is likely to lead to cover for biased editing and problem editing and cause many more, and different, problems than what Sanger suggests. It should be removed because it has gone well beyond a proper purpose of "there are exceptions to every rule" and even beyond reasonable alternative interpretations can at least be discussed. Perhaps an essay or some more moderate and properly explained guideline could be added in its place.
Exposing real names of "leaders" and contributors will inevitably lead to doxxing, threats and other unwanted and unneeded exposure. Editors and users will leave Wikipedia in droves. As much trouble as biased, incompetent, vandalistic and other types of bad, troublesome or incompetent actors are, and as much time as they waste being corrected and often eventually blocked, they are a lesser bad option that needs to be combatted than the enormous drop in edits and contributions that would occur (maybe not in sports and pop culture, but just about everything else). How many editors does Citizendium have and how fast has it grown? How many views does it get? Too few, I would venture, to be held up as a model for Wikipedia in most respects. FWIW. Donner60 (talk) 01:45, 4 October 2025 (UTC)[reply]
From what I gather from the interview, Sanger is suggesting the naming of Arbitration Committee (Wikipedia) and the Checkusers not the admins or regular users/editors. He says that he is against doxxing and that if such policies are adopted, those with such rights should be allowed to voluntarily retire if they want to stick with anonyminity. This, as I elborate above, appears to be reasonable. Gotitbro (talk) 06:36, 4 October 2025 (UTC)[reply]
They can all retire, and no one would dare take their places for fear of being doxxed or threatened. The fact that Sanger is against doxxing won't prevent doxxers from doxxing and certainly won't prevent threats, or worse, against revealed editors. Taking away the arbitration committee and the checkusers, which is what revealing their names would do, would hamper controlling and eliminating bad actors, POV pushers (especially skilled ones), trolls and vandals and prevent solving the most troublesome old and new problems they would create. These bad actors will be first in line to be doxxers. Eliminating such control over serious problems and people, which supporters of this position need to realize will happen, as it does in other areas of life, will degrade Wikipedia. I see no good reason and no benefit from disclosing these names and the resultant damage to Wikipedia which would inevitably take place. Are there instances that can be pointed out where arbitrators and checkusers have caused any sort of damage to the project, instead of obvious benefits to the project which have resulted from their actions? Donner60 (talk) 06:52, 4 October 2025 (UTC)[reply]
I see your point, but threats of doxxing only work when you don't have public faces. As I state above this could indeed engender more community and public trust. Resilience against MAGA and similar movements worldwide shouldn't hamper our own internal initiatives. Bad-faith actors are extant regardless. Though rather than online trolls et. al. what might actually precaution such moves is the existence of events such as List of people imprisoned for editing Wikipedia. Gotitbro (talk) 20:59, 4 October 2025 (UTC)[reply]
Your last point is a good one. We are in at least partial agreement, if not in entire agreement because of your last point. Threats could be made, even if not technically doxxing, if names are revealed. I have even been threatened at least twice due to edits on Wikipedia. I think checkusers tracked them down and said they were empty threats because of where they were located. We have stated our opinions and there is little, if any, reason for us to continue to discuss this point. (Not trying to preclude you from another comment but I may not further reply.) Thanks. Donner60 (talk) 22:37, 4 October 2025 (UTC)[reply]
I don't think there's a good reason to single out CUs or crats. The stated reason is control over content. If I or any other editor familiar with the PAG got into a content dispute with an arb, CU or crat (that did not end up resolved after the standard BRD) I have no reason to believe that they're more likely to prevail than I am in a 3O or an RfC or whatever. If we want accountability of content, I think we should invite comment from academics in a field. Maybe get the WMF to pay some or something. How much are journal editors paid? It might be worth having verified and/or public qualifications or identities for those. Even if the community probably wouldn't want to cede the editorial process to editors the WMF hires, I'm sure we'd be open to some degree of expert involvement. Alpha3031 (tc) 18:22, 5 October 2025 (UTC)[reply]



       

The Signpost · written by many · served by Sinepost V0.9 · 🄯 CC-BY-SA 4.0