The Signpost

From the editor

What else can we say?

Last week the Arbitration Committee made an unusual decision involving the policy on outing. An editor was topic-banned based on an article published in The Daily Dot which apparently outed the long time editor. The unusual part was that ArbCom specifically allowed editors to link to that article in discussions about conflict-of-interest editing. Some editors were delighted that an offender of our COI rules had been caught. Others fear that ArbCom was unilaterally changing a major protection against harassment. A discussion at WT:Harassment seems to indicate that the editors there don't like the idea of allowing linking to a reliable source that both names an editor's real name and their username. But it looks like they'll accept links to the Daily Dot article now since ArbCom has given special permission. The Signpost has a different take on the controversy.

We've been trying for years to report serious COI and paid editing problems while staying within the outing policy. We believe that it is outrageous that accused sex offenders like Jeffrey Epstein, Ghislaine Maxwell, and Peter Nygard were apparently able to edit Wikipedia, either as COI editors or through paid editors, though we recognize that administrators were successful in catching or at least slowing down these editors.

We linked to The New York Times where they gave a username that they state was named by Epstein in an official document as an account that he used. Since we also had on-Wiki evidence of paid editing including the following edit summary, we believe we were following the current outing policy.

Edit comment by User:Stgeorge12

To be clear, the policy, as written, is convoluted, at times self-contradictory, and just poorly written. Violating the policy has dire consequences – up to an immediate ban that can be imposed by any administrator. We just wish that the community would write an easy-to-understand policy that would apply to everybody. We don't want to ask ArbCom for exemptions for specific stories, which would look like submitting our stories to pre-publication censorship.

Last month we reported that accused sex offender Peter Nygard apparently used paid editors to edit Wikipedia. The evidence included a story in The Guardian, which we linked to, that documented a UK court order requiring that the Wikimedia Foundation release the identities of editors. The court order could not be enforced, but there was strong evidence on-Wiki that Nygard, or more likely, his paid representatives, edited Wikipedia.

The essence of the outing policy is that an editor's username should not be connected to "personal information [including] legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph, whether such information is accurate or not," unless the editor has stated that information themself on the English Wikipedia.

When discussing COI or paid editing "workplace address, job title and work organisation", can be tricky but we believe that simply establishing a strong possibility that the editor behind a username is a paid editor does not out their workplace address or work organization. For example a COI or paid editor may work for an individual person, a company that is the subject of an article, or their lawyer, or a paid editing company. Or they may just be a friend or relative editing from their own home or workplace.

We request that the community – and ArbCom if they wish – take another look at some of the articles we've written in this series, and let us know what, if anything, they think violates the letter or spirit of our policy on outing. Then the community will have an idea of what needs to be clarified – or possibly changed – in the policy for us and other editors to be able to seriously discuss possible paid or COI editors inserting disinformation into Wikipedia. Not having a clear policy on outing will prevent Wikipedia from addressing these matters in a world full of disinformation.

Please review the following Signpost articles and let us know in the comments section below whether you believe we have violated the letter or spirit of Wikipedia's outing policy

Please note that we are not asking for a special rule for The Signpost. Outing policy should apply to all editors, whether they are editing at WP:COIN, at a sockpuppet investigation, on an article talkpage, and even in mainspace articles.

Given a clear rule The Signpost can continue to report on COI and paid editing, albeit in a somewhat less informative way if the rule is extremely strict. Really? What if the policy prohibited mentioning any real-world names, usernames, anything related to a specific workplace or even location, as well as the current prohibitions?

The following section presents a story, using the above super-strict rules, that we were going to report in the normal way in this issue. Please let us know in the comments section whether you think that a story presented in this way is useful and informative, or whether it violates the dignity of a fellow human being and shouldn't be allowed even under such strict rules.

The case of the boss's father's article

A high-level government official who is married to a very well-known legislator was recently reported to have ordered government employees on January 6, 2018, to edit the official's father's Wikipedia article. The report was made by a high-level investigating office who forwarded it first to prosecutors, and then to the national legislature, who then released it to the public. At least two reliable national news organizations then reported the Wikipedia connection. Two news sources posted the slightly redacted investigative report online.

The government employees told the investigators that their boss had told them to perform many possibly illegal actions, with editing Wikipedia – which is just an infraction of WMF's terms-of-service ban on undeclared paid editing – a minor part of the report. An employee told the investigator that editing their boss's father's Wikipedia article was legal because it was important for the boss to present themself to the public as a dutiful son or daughter. Almost like clock work, an editor made a minor edit, their first and last edit, to the rarely edited article about the father on January 6, 2018.

From there the case of the boss's father's article gets a bit murky. While there were multiple new editors cleaning up the article from time to time, there's only a small amount of evidence that these editors were sockpuppets or had been banned for conflict-of-interest editing. Most of the edits were related to a slow moving edit war involving naming conventions in a country other than the one where the high government official served.

Similarly the article on the boss's mother, and the article on the boss's father's company showed only a few signs of conflict-of-interest editing.

What is clear is that the article about the boss was heavily edited by apparent COI editors, who were likely undeclared paid editors. Long, fairly unimportant material that appeared to be government press releases about the boss were inserted a dozen times by single-purpose accounts. This press release material was reverted by other Wikipedians as often as it was added. The single-purpose editors were questioned about their editing but did not respond and simply disappeared. About twenty accounts were blocked as sockpuppets, meatpuppets or COI editors; a couple of them had also edited the father's or mother's articles.

We're asking you

The apparent COI editing in this story has been discussed on talk pages, WP:COIN, and in a sockpuppet investigation. All of these discussions at least implicitly referred to the high official's name and their spouse's name, the government agency where they formerly worked, as well as the COI or paid editors' likely place of work in the government.

Is there any reason that this publicly available information can't be brought to a wider audience at The Signpost or elsewhere on Wikipedia? Is there any benefit to anybody if this publicly available information couldn't be reported here or elsewhere on Wikipedia? Is there any substantive reason that a reliable media source or the official investigator's report couldn't be linked to document the facts?

What is the cost to Wikipedia and its readers of not reporting this information?

Please just let us know in the comments section below.

What else can we say?


+ 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 know the subject in question that is being alluded to here (I'm the one who raised the matter with The Signpost I believe). Was there an official threat of banning or something that has led to this? Where's the editorial discussion that brought about this convoluted way of discussing the problem? -Indy beetle (talk) 05:40, 29 March 2021 (UTC)[reply]

    • I think 3 folks sent me that story, so I'd have had it anyway. You should know that I'm fairly skittish on actually publishing these stories. They need to be done IMHO, but everything has to be just right, quadruple checked, etc. to publish. If I feel that I don't understand how people are interpreting wp:outing this week I may not publish at all. i intend to keep doing these stories. It's not that I'm a coward if I don't publish if there's a 5% chance of getting banned, It's that I intend to do 12 plus stories on this general topic this year, and it only takes one story a little bit off to get permanently banned. How can I get to 12 if I get banned after the 3rd one? I take full credit (and blamr) for these stories, but it was approved by another editor. Smallbones(smalltalk) 02:23, 30 March 2021 (UTC)[reply]
  • If Signpost ever are forced to write in "The case of the boss's father's article" style then make up some fake names to increase understanding and emotional engagement in the story. Every time I consider the tension between WP:OUT and WP:COI (or identifying other issues like off-wiki brigading and harassment) I reach the conclusion "there is no solution to this tension". But no answer on the matter is not an option. Perhaps a well-planned RfC could generate a worthwhile discussion. I think the Signpost's reporting, including the four linked articles, is valuable, but what's less obvious is that there is a consistent approach to outing which allows such writing but disallows privacy violations as much as possible. — Bilorv (talk) 10:03, 29 March 2021 (UTC)[reply]

WP:OUTING as written is a childish piece of policy that is an embarrassment to what, after twenty years, should be a mature project. It is unbalanced. It is reminiscent of the "libel laws" practised in dictatorships, where you end up in jail when you dare criticise a functionary. It is not fit for purpose. It's been regularly abused by the likes of Tenebrae and Wifione (for the real-world damage the latter admin did, see Newsweek) – all they had to do was shout "harassment" and "outing", and trigger-happy admins would threaten, block and ban people drawing attention to their editing.

The policy's chilling effect is evident from the fact that in Tenebrae's case, someone created an SPA to raise what were – with the benefit of hindsight – clearly justified concerns, rather than risking their real account. Can we really blame them, when WP:OUTING calls for an "immediate block" (rather than suppressing the edit and advising the editor on what channels to use instead) for any post touching even remotely on an editor's identity, regardless of whether there is a good-faith concern about the integrity of the encyclopedia or not?

All the SPA in Tenebrae's case posted was diffs, as one would in an ArbCom case, yet they were still blocked in response to Tenebrae's cry of "harassment and attempted outing", which was immediately endorsed by Baseball Bugs and Knowledgekid87 and swiftly responded to with a block, all done in less than two hours. Quick work! As it was, another six years passed before the issue was raised again – off-wiki, i.e. in the press – and appropriate action was taken on-wiki.

Drmies, are you not embarrassed by this block today? It's kind of telling that the two administrators who blocked the Nola Carveth SPA and suppressed the discussion of Tenebrae’s COI editing (now thankfully unsuppressed again by Beeblebrox) are among those who object most loudly to the sort of bare-faced lying that Wifione and Tenebrae (or Qworty, for that matter) got away with for so long. In other words, even admins who intellectually understand the problem still are drawn to do the wrong thing when actually faced with such a situation, as though on autopilot – through peer pressure, I guess. What is the solution? --Andreas JN466 15:14, 29 March 2021 (UTC)[reply]

  • Oh, I don't know if there is a solution (banning Tenebrae isn't a solution to a systemic problem, though it is the right thing to do), but what that editor was doing was going about something in completely the wrong way. Hindsight is 20/20, but dropping a brief but accusatory nugget out of a much larger matter on the most visible talk page on the project, that's not helpful. If that editor can find their way to Jimbo's talk page, I would expect them to be able to email ArbCom--or, in this case, the administrator who blocked them. And if they had done the latter, this administrator would have looked into it. But I have kind of an aversion to shit storms that come in out of nowhere. Six years passed--the editor could have emailed me or anyone else a thousand times in those years. Which, ironically, would have greatly reduced the harm done by the COI editor. Drmies (talk) 15:47, 29 March 2021 (UTC)[reply]
    • I am involuntarily reminded of Li Wenliang, who was accused of "making false comments" that had "severely disturbed the social order". I am sure someone told him, too, that he just went about reporting the problem the wrong way. Really, I cannot see much difference. --Andreas JN466 17:29, 29 March 2021 (UTC)[reply]
      • That's all you, this version of Godwin's law. Let's be clear: your Nora is an experienced editor who knew what the better way was. Drmies (talk) 03:13, 30 March 2021 (UTC)[reply]
        • Judging by commentary on WO, their "experience" may have told them that the "better way" is a mirage, because there is no sign of it working. That seems to be the consensus over there. As I see it, you were simply played, just like admins defending Wifione were. Tenebrae cried "harassment and attempted outing". Feeling your "aversion to shit storms" triggered, you obliged him instantly. You went for the easy satisfaction of insta-blocking a throwaway account and had a nice pat on the back from Tenebrae for getting everyone off his back, along with some flattery and a lecture from him on what a great editor he is. And six years later, you're left telling Tenebrae how "pissed" you are because he deceived you. And yet you think you did everything right, and would act the same way again today. --Andreas JN466 08:29, 30 March 2021 (UTC)[reply]
        • If you're not prepared to read WO, I believe AndyTheGrump, among others, has a more comprehensive grasp than me of how the "better way" failed in this case. Cheers, --Andreas JN466 09:36, 30 March 2021 (UTC)[reply]
  • I think Smallbones raises some very good points re the tension between outing and paid editing. I have mixed feelings about Tenebrae, as I've pointed out elsewhere. Yes, hindsight is 20-20, but here is what I don't understand about the approach taken by the indeffed editor Nola Carveth. The appraoch they took was aimed at the COI, when simultaneously the McDonagh and Lovece articles were a mess. I know that without going back too far in the page history because they were still a mess a few days ago. I'd have thought that focusing on the terrible shape of those articles would have been more effective than the path taken by that account. Remember that there are several systemic problems at work here. One of them is that bad articles that are little more than resumes are more common than not. Another is that dealing with those bad articles, especially when intent paid and/or COI editors and socks are buzzing around, is a drain on editors. It is time-consuming, sometimes risky, and demoralizing. Coretheapple (talk) 20:54, 29 March 2021 (UTC)[reply]
    • Didn't you answer your own question at the end of your post? "Dealing with those bad articles, especially when intent paid and/or COI editors and socks are buzzing around ... is time-consuming, sometimes risky, and demoralising." Tenebrae was prolific, and his COI affected hundreds of articles, judging by the number of Wikipedia articles (about 650) mentioning/referencing Lovece or his wife today. I don't think anyone will ever bother ... each mention is probably no big deal when viewed in isolation, and those articles that are likely to have strong NPOV issues seem to be fairly obscure, fancruft almost, with only a small number of page views per day. My best guess is Nola Carveth wanted to illustrate the systemic problem in a prominent place where it would get eyes. If the problem got solved, as things sometimes did on Wales's page in those days, good; if not, then a blocked SPA sock was no big loss, and at least the system would fail publicly, and be seen to have failed. That wouldn't be my style, but I can't deny it would have its own logic to it, and even a kind of integrity and good faith, given that there was an actual problem that needed addressing, and that whoever raised it would inevitably put themselves at risk of being blocked. --Andreas JN466 22:37, 29 March 2021 (UTC)[reply]
      • By the way, imagine for a moment if Tenebrae had played the Wikipedia game with greater finesse, had disclosed his identity at the outset and been less aggressive. He'd have made COI a non-issue and I suspect that the articles in question would be much the same as they were before this fiasco. Coretheapple (talk) 13:30, 30 March 2021 (UTC)[reply]
      • As if to further make the above point, a well-known organization, ADL, has been engaged in so-called "white hat" COI editing on various articles throughout the project, and just posted on the COI noticeboard that, in effect, it will continue to do so and if we don't like it we can pound sand. That brings to mind how so-called white-hat COI editing can be just as much of a time suck as the "black hat" variety that is a drain on community resources. Coretheapple (talk) 18:54, 30 March 2021 (UTC)[reply]
  • When I was a member of arb com , after several years of trying, I with the help of other arbs persuaded the committee after several year of trying, to accept some responsibility for enforcement of the terms of use, particularly regarding paid coi. The argument that was successful in this was exactly he need for dealing with confidential information--it was decided that this would be handled primarily by the checkusers, as this was the likely confidential part of the data. Since then the committee itself has become more involved, and I am glad to see it. It's the only fair way of resolving the dilemmma between privacy and coi. I think it is be clear that the case leading to the article was a particular difficult one, and I remain uncertain whether the correct decision was made. I have the impression that others involved may feel similarly. If humans deal with matters involving other humans, they will make mistakes. Working jointly , ascheckusers do in difficult cases and arb com always does, the mistakes will be less frequent than if a single admin handles matters, but they will still happen. DGG ( talk ) 04:06, 2 April 2021 (UTC)[reply]
  • I made a separate paragraph for this, because having worked with them, I think our current rules untenable. The conflict between privacy and coi enforcement is far too frequent and the results of our processes are in this sort of situation far too unpredictable . The people who have commented that there is no safe way for an admin or other editor to proceed are correct. The net effect is that we are all generally more likely to deal with coi by guesswork than information, which is not a satisfactory way to resolve the dilemma.
I suggest that we say, as the WMF rules on privacy I think permit us to say, that in order to protect the encyclopedia, editors who are not editing in good faith cannot expect to have their privacy fully protected, and that conflict of rules be resolved in favor of protecting the encyclopedia. This is explicitly stated with respect to sockpuppettry; I suggest this may need to be specifically stated witth respect to COI. . (I'm not making a formal proposal here, and am therefore not suggesting what we and the Foundation might consider acceptable specific wording, and I am not giving references, because the wording is spread out over a considerable number of pages on various parts of the site.) And of course, I personally will continue to act in the most conservative way according to the generally accepted interpretation, to the extent I can figure it out. . DGG ( talk ) 04:06, 2 April 2021 (UTC)[reply]
  • I think the first step is to turn WP:COI into policy. This might dissuade vested editors and administrators from flaunting their COIs and daring mere mortals from doing anything about it, as I have personally observed in the past. Coretheapple (talk) 20:54, 18 April 2021 (UTC)[reply]

















Wikipedia:Wikipedia Signpost/2021-03-28/From_the_editor