Wikipedia talk:Arbitration/Requests/Case/Conflict of interest management/Workshop

Oversight vs. Involved
Since we quibble over quibbles in policy discussions, I note that INVOLVED does not mention functionaries. It talks about admins first, and at the end says the same principle, not policy, applies to discussion closers. The oversight policy does not talk about Involvement. Oversight isn't even an admin tool. It is our practice, not policy, to not give non-admins oversight. Involved does not apply to non-admin tools. On the flip side, we don't have anything that says ovesightable information is an emergency situation. There are no special allowances made for oversight blocks either. It is only a template and the only policy it prescribes is that the block may not be adjusted or vacated except by oversight team and arbcom. Ultimately, I see no policy basis for most things that are taken for granted in this area. Not everything needs to be written, or can be written, but when we have no written guidance on matters that have become controversial, every position becomes defensible. — Usedtobecool ☎️ 12:23, 16 March 2024 (UTC)


 * From my 17 years experience, I would say that WP:INVOLVED is really applied to ANY advanced tool, from rollbacker to OS/CU, to one degree or another. History has shown us that it is expected that someone won't use any advanced tool if they are WP:involved, except for limited circumstances (vandalism, BLP, etc)  Whatever the community consensus is (community norms), that is the real policy.  The written stuff is just a convenient documentation of it.  I know, this isn't necessarily intuitive, but it is how Wikipedia works.  Maybe the written policy needs to catch up, but it is still considered policy, and least from my read of consensus over the years.  Dennis Brown - 2&cent; 03:36, 28 March 2024 (UTC)

Anyone remember?
This comment moved from in reply to HJ Mitchell's question to Nihonjoe, as El C is not a party nor was he who the question was directed towards. Barkeep49 (talk) 22:34, 21 March 2024 (UTC) Un/related, who here remembers the WP:AMA? El_C 20:45, 21 March 2024 (UTC)
 * Good advocacy. (For a fair and unbiased proceeding, of course.) El_C 22:59, 21 March 2024 (UTC)

Principles to cover
I'm wondering if anyone wants to propose any principles about: I'd also be interested to hear any more thoughts about oversight blocks vs INVOLVED.
 * Admissibility of evidence from other Wikimedia projects as it relates to OUTING
 * The standards of conduct expected of bureaucrats
 * The tension between OUTING and COI (and/or how it can/should be resolved)
 * Whether editing with an undisclosed COI is compatible with adminship/'cratship
 * Whether any of Nihonjoe's edits were problematic in their own right
 * How much information an editor with a COI needs to disclose (ie, is it enough to say "I have a COI" or use connected contributor or do we need to know more about the relationship?)

I considered proposing some principles myself but I'm actually more keen to hear opinions from within the community. HJ Mitchell &#124; Penny for your thoughts? 16:42, 25 March 2024 (UTC)


 * Did #2 and #4 if that helps. Noah, AATalk 18:15, 25 March 2024 (UTC)
 * Let me know if those help. Noah, AATalk 18:32, 25 March 2024 (UTC)
 * Did #1 and #6 too. I did two points; one for regular COI and one for PAID. Noah, AATalk 23:04, 25 March 2024 (UTC)
 * Anyone considering a principle regarding #1 should probably read . Sincerely, Dilettante 18:23, 25 March 2024 (UTC)
 * Don't worry, a proposed principle doesn't trump an RfC but these are some of the issues that arbs are going to have to grapple with so thoughtful proposals on how (or, indeed, whether) we should handle them can't possibly hurt. <b style="color: teal; font-family: Tahoma">HJ Mitchell</b> &#124; <span style="color: navy; font-family: Times New Roman" title="(Talk page)">Penny for your thoughts? 18:49, 25 March 2024 (UTC)
 * Sorry if my phrasing was unclear. I meant that any proposed principle should take the discussion and close into account because they raise various arguments in favor and against of various possible proposals, not that I'm worried about consensus being overridden. Sincerely, Dilettante 19:47, 25 March 2024 (UTC)
 * I hadn't read the RfC until after I already wrote the principle for #1, but that reflects what I would consider fair game. What are your thoughts (point #10 under my section)? <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 23:16, 25 March 2024 (UTC)
 * , I think it's probably the most reasonable interpretation of OUTING and well-worded too. Sincerely, Dilettante 15:27, 26 March 2024 (UTC)

Alternate Principles on Undisclosed COI
User:AKAF has proposed a finding of fact that Nihonjoe had a good reason for not disclosing a conflict of interest, which was in order to remain pseudonymous. Their proposal has had two different interpretations, both of which were reasonable interpretations of the words, but one has an unreasonable implication, and I think that it is important to discuss the two alternate interpretations. The first is that an editor may avoid disclosing a COI in order to protect their pseudonymous status, and so may make undisclosed COI edits. The second is that an editor may avoid disclosing a COI in order to protect their pseudonymous status, but must completely avoid any edits in any way affecting that COI. The first is a completely terrible idea, because it would allow spammers to conceal their vested interests so as to preserve their pseudonymity. That seems to be what AKAF wrote, but we can assume good faith and assume that is not what they meant.

However, there is also a proposed principle by User:Hurricane Noah that undisclosed conflicts of interest are incompatible with the trust of an administrator and a bureaucrat. That principle would mean that administrators who use pseudonyms must disclose their conflicts of interest anyway, so that the community can maintain trust. The ArbCom does need to decide whether undisclosed COIs by pseudonymous admins are permitted as long as the pseudonymous admin is careful to avoid any COI edits.

So there are three possible principles on COI by pseudonymous admins, one of which can be discarded:
 * 1. Undisclosed COIs by pseudonymous admins undermine trust.  COIs by admins must be disclosed, and pseudonymous status is not a reason to avoid disclosure.
 * 2. Undisclosed COIs by pseudonymous admins are permitted, only if the admin takes great care to avoid editing where a COI is present.  The community places trust in the admin to avoid edits where there is an undisclosed COI.
 * 3. Undisclosed COIs by pseudonymous admins are permitted, and maintaining pseudonymity is a reason not to disclose.

So ArbCom should decide whether it is 1 or 2. Robert McClenon (talk) 21:58, 25 March 2024 (UTC)
 * Pinging User:Valereee, who also commented on this discussion in the workshop. Robert McClenon (talk) 22:01, 25 March 2024 (UTC)
 * I followed up with points 8 and 9 to define what would have to be shared. Does that help a bit? <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 23:02, 25 March 2024 (UTC)
 * User:Hurricane Noah - You answered questions that I didn't ask. Points 8 and 9 have to do with COI edits in general, and don't appear to be about COI edits by pseudonymous administrators.
 * We agree that all COI edits must be declared. The question is whether admins must disclose COIs when they are not editing with regard to them.  Robert McClenon (talk) 02:25, 26 March 2024 (UTC)
 * I'm not sure if it has been brought up in the past or not, but the way I see it, any conflicts that could potentially sway the impartiality of an admin in situations should be disclosed. My one point made it clear that a person simply stating they have a conflict with an article would be enough, unless it is paid editing. The bar for when to share is the only thing that is stricter for admins according to my principle specifically mentioning them. <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 02:38, 26 March 2024 (UTC)
 * I think it's commonsensical that admins and other high-tier advanced perms holders would need to disclose CoIs to prevent the possibility of them engaging in admin actions or closes with undisclosed CoIs.
 * I don’t believe this should be enforced harshly, though, because many people have more potential CoIs than they could reasonably be expected to list.
 * In terms of how to enforce this at all, except for the cases when people are clueless about covering tracks (as here), or how to avoid another Wifione/Lourdes-style infiltration, well…
 * The only idea that comes to mind is perhaps making crats subject to the same or similar verification standards as functionaries.
 * RadioactiveBoulevardier (talk) 03:39, 26 March 2024 (UTC)
 * I disagree that people should be required to disclose COIs that are irrelevant to their editing or other actions they take on Wikipedia. To do so would require many people to out themselves, potentially placing themselves at the risk of real-world harm, with no benefit to the encyclopaedia. For example, if User:Example is both an admin a senior executive of a notable laboratory that tests cosmetics on animals they obviously have a conflict of interest regarding that laboratory, animal testing and closely related articles. However if their editing and admin activities do not include that area (say they focus primarily on textile arts) then their COI is irrelevant to Wikipedia and we gain nothing by knowing who they are in real life. Thryduulf (talk) 04:04, 26 March 2024 (UTC)
 * I certainly didn't wish to encourage COI editing (thanks for assuming my good faith), but I wished to encourage a finding about the (sometimes difficult) tension between acknowledging COI and remaining pseudonymous.AKAF (talk) 11:49, 26 March 2024 (UTC)
 * How I see it is that a problem can always arise if they are asked to take admin actions in areas where they have an undisclosed conflict. This can be very problematic of course. What I am saying would be required is simply stating one has a COI with an article. They need not describe the relationship unless they are being paid to perform edits which of course would require them to do so. <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 12:11, 26 March 2024 (UTC)
 * You are never* required to use your admin tools in a given situation, even if asked. If admin has a COI regarding e.g. animal testing that (for whatever reason) they do not wish to disclose and get asked to take action in that area then they have three possible options:
 * If and only if the action is urgent, obvious and any other admin would take the same action, then they can take that action with no issues. For example protecting a page on which there is an active edit war or blocking an active spammer.
 * Politely decline, with or without explaining why, e.g. "I'd rather not get involved in that topic area, sorry".
 * Wait until after some other admin has acted before responding so the request is moot.
 * Obviously if there is any doubt about whether an action is both urgent and obvious then only options 2 or 3 should be considered. Option 1 is within the same spirit that allows even PAID editors who should not normally be editing the page directly to revert obvious vandalism.
 * * well, arguably there are some occasions when you should if you are asked to self-revert or take follow-up admin actions, but these are not relevant here. Thryduulf (talk) 12:32, 26 March 2024 (UTC)
 * Here's the problem with the "this outs them, so we should never ask!" tack: Nihonjoe already functionally outed himself. Any longtime editor who has any variety of interests has created a substantially-unique, publicly available corpus of their writing style, interests, likely geographic or national origins, etc. So much of the OUTING policy is fundamentally at odds with how Wikipedia actually works, how people connect their accounts across the web, and that it's not up to the community to cover for your own actions (where I've seen people interpret OUTING so stringently that even telling people your name on Commons or another WMF site somehow means that they haven't outed themselves on en.WP. Which, I mean, come on.
 * We are seeing with this case what happens when an admin doesn't disclose a COI in an area they theoretically weren't editing (by their own estimation, which I think the evidence amply proves was seriously flawed)—they ultimately very much were influencing that topic, and no one had any clue that had any connection (and once that connection was brought to light, the fact it was never disclosed makes it much harder to buy an innocent connection.) The lure to use anonymity as a shield against scrutiny is always going to be tempting to ignore. At the very least, we should expect better from a bureaucrat and admin rather than a regular user without advanced permissions and ability to more materially sway content and conduct. Der Wohltemperierte Fuchs  talk 12:44, 26 March 2024 (UTC)
 * If is in response to my comment, then you have misunderstood what I'm saying. I'm not saying that it is always inappropriate to ask whether someone has a COI (although there are times and ways it can be), I'm saying it is inappropriate to preemptively require admins to disclose COIs they have regarding areas they do not edit/admin in. This not relevant to situations where an editor is editing/admining in a given area despite claiming not to. Thryduulf (talk) 13:08, 26 March 2024 (UTC)
 * ...But why not? This isn't even talking about a random editor, this is someone being entrusted with special permissions by the community. I don't think "hey do you have any areas or topics you avoid because of a COI" is invasive or dangerous, and it's frankly a lot easier to deal with misconduct if they're transparent (or lying about it) from jump. Der Wohltemperierte Fuchs  talk 19:01, 26 March 2024 (UTC)
 * Because of the chilling effects of requiring admins to out themselves, potentially with risk of real world harm, with no benefit to the project. Yes, it's easier to detect if they're lying, but we don't write policy on the basis of trusted users lying to us. Thryduulf (talk) 19:13, 26 March 2024 (UTC)
 * We aren't saying to disclose the nature of a relationship. One could have a COI for many reasons, even simply I despise X person or y organization so much that it would affect their impartiality. Broadly stating one has a conflict without anything further is different from someone outing themself. <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 19:25, 26 March 2024 (UTC)
 * Saying "I have a COI regarding MegaCorp" doesn't tell us anything useful - it could be they are the CEO, it could be their partner is a part-time caretaker at one branch. And you know as well as I do that people (on and especially off-wiki) will attempt to find details if they are involved in anything remotely controversial (even if entirely unrelated) or they simply dislike like them. Secondly, what benefit do we get from knowing that an admin has a COI with a topic they don't touch? If they are going to lie about whether they have a COI, they will just omit to mention it during mandatory disclosure. Thryduulf (talk) 19:36, 26 March 2024 (UTC)
 * It does say that a conflict exists for them, which is useful enough on its own. The benefit we gain is knowing that our administrators and bureaucrats are not editing topics with which they have a conflict of interest, given they are transparent. A case like this unfortunately highlights the need for better enforcement. <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 10:33, 27 March 2024 (UTC)
 * Except we don't gain that knowledge. If I had a COI with MegaCorp but wanted to edit that without disclosing that, I could just list COIs with a couple of other topics that I have no intention of editing (whether I actually have a COI with them or not) and omit to mention MegaCorp giving the illusion of transparency. Of course those acting in good faith wouldn't do that, but then those acting in good faith wouldn't edit where they had a COI without the disclose either. So it compromises privacy for at best no gain. Thryduulf (talk) 11:47, 27 March 2024 (UTC)
 * Fair enough. I will make a change to it before the workshop closes. <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 23:08, 27 March 2024 (UTC)

Definitional problem with "conflict of interest"
Since the workshop hasn't closed yet, making a section here on my proposed principle "Conflict of interest includes but is not limited to financial conflict of interest."

This has apparently gone over like a lead balloon, but if COI that doesn't involve money is either seen as unimportant or not classed as a COI at all, then aspects of this case, and the cases of many other editors who've been accused of COI editing on noticeboards and at ArbCom, will be judged very differently. (e.g.: Using self-authored online documents as major references; Creating articles about family members.) If the only COI we really care about is COI involving money, then there will be some degree of difference in how we judge editors who don't work in business. Different viewpoints about non-financial COI have been a source of disagreement before and during this case, so to contextualise whatever the Arbs decide, I do think it's necessary for them to speak to how seriously we take COI that cannot be called paid editing. Yngvadottir (talk) 09:51, 29 March 2024 (UTC)


 * I see COI as as a spectrum rather than an absolute state. We all have personal/professional/financial interests outside of Wikipedia and it's inevitable that people will write about what they're interested in but some outside interests are more concerning than others when it comes to writing a neutral encyclopaedia. A legal relationship is quite easy define and is definitely a case where a COI should be disclosed if you're making substantive edits to the subject's article but it arguably extends all the way to people who want to tell the world about their home town. <b style="color: teal; font-family: Tahoma">HJ Mitchell</b> &#124; <span style="color: navy; font-family: Times New Roman" title="(Talk page)">Penny for your thoughts? 16:57, 29 March 2024 (UTC)


 * As I've said elsewhere, I see it as a continuum; there are many reasons why someone may have a COI, as well as many degrees of COI. It contributes to unconscious bias as well as to deliberate POV editing. Religious background is an example; apart from obvious issues like preferences in sourcing and more or less deliberate formation of like-minded editor groups, someone may have a whole raft of assumptions about what's notable, what should be emphasised, how things should be defined ... and what's not notable, what's trivial, classification, and inclusion. Yngvadottir (talk) 21:25, 29 March 2024 (UTC) Adding after some coffee: However, the "unconscious bias" parts fall outside the "COI that should be declared", or we wind up advocating editors not only not be anonymous, but make comprehensive declarations on their user pages about their backgrounds, affiliations, and philosophies. Rather, we rely on editorial give and take, and for that to work, we need to recognise that we all have both conflicts of interest and unconscious biases, and that these extend to coming to the ideas of conflict of interest and of bias with differing blind spots. Wikipedia is a collective effort, not an enterprise with assigned tasks. The editorial process works when other editors provide corrective influences; beyond article editing and deletion discussions, the same principle extends to the treatment of POV pushers and to discussion with fellow editors. But limiting COI to financial COI is itself a bias based on personal experience and assumptions. IMO. Yngvadottir (talk) 21:58, 29 March 2024 (UTC)

Animal lover 666's comment
Could an arb or clerk please sign the most recent comment made by Animal lover 666? Didnt want to do so since editing is forbidden at this point. <b style="white-space:nowrap;text-shadow:#009200 0.3em 0.4em 1.0em,#009200 -0.2em -0.2em 1.0em;color:#009200">Noah</b>, AA<b style="color:#ff0000">Talk</b> 16:59, 29 March 2024 (UTC)

Intersection of COI, INVOLVED, and urgent priorities
Before realizing that the workshop is now closed, I reflected on 's trial principles and thought the idea behind them, as well as some of the issues with several parties' actions in this case, might be better served by a principle like the following. If this is out of order and I'm late to the party, so be it, but pasting here in case someone finds it useful. It would put the onus on advanced permissions holders how to meet expectations of transparency and ensure review of urgent noncompliant actions rather than trying to prescribe/describe exactly what WP processes, like oversight, have or don't have adequate review.


 * Given the potential to introduce pernicious bias, the Wikipedia community is strongly concerned about, and has paid increasing attention over the years to COI editing and involved administrative actions. Nevertheless, in spite of their high importance, WP:COI and WP:INVOLVED have remained as guidelines, subject to exceptions and with only selected elements codified as strict policy, in part because urgent priorities, for instance related to WP:OUTING or WP:BLP, may on occasion be even more important. However, the community expects that users with advanced permissions in particular will be highly transparent when such instances arise. In situations with limited urgency, they should seek other solutions than to make noncompliant edits or actions themselves. Where urgency requires violating these guidelines, they should proactively ensure that their edits/actions are reviewed by an unconflicted/uninvolved 3rd party.

Martinp (talk) 20:30, 29 March 2024 (UTC)