User:Nosebagbear/UPE Proposals

The following are a list of potential proposals on handling undisclosed paid editing, that I've grouped into 3 categories: coverage, restrictions/limitations, and enforcement. I've tried to include a brief neutral summary for each. The idea is to send forward ideas that aren't bonkers for Community discussion and feedback, rather than a single combined policy that lives or dies as one. Obviously, like with the admin activity RfA, contradictory proposals exist and would default to the one with the most support.

Some proposals would need their own RfC if supported to designate all the specifics. This is needed to avoid just the proposals part of the discussion running for 50 pages.

[For idea discussion here and ultimately RfC !voting] If you believe that a proposal should be passed but only with relevance to a subset of "paid editors", please mark that prominently in your !vote.

1. Please add your own proposals in the appropriate category
2. If my categories seem incomplete, please feel free to add your own. If they seem wrong (or proposals should belong in other categories), please consider first discussing on the Talk Page to avoid confusion.

Nosebagbear (talk)

Coverage

This category includes who paid editing policy should apply to, or if a subset could be formed to have further restrictions applied to them.

1.1) Remove paid editing limitations

This would move us in line with other communities that have opted out of paid-editing disclosure policies, with paid editors free to participate under exactly the same conditions as regular editors.

Discussion of the idea

This is very unlikely to happen, and I don't think it would be a productive use of discussion resources. Seraphimblade ^{Talk to me} 15:36, 21 July 2020 (UTC)

I personally agree, and think it's probably the least likely to warrant including in an ultimate RfC, but I put it in here as individuals do sometimes suggest it. Nosebagbear (talk)

1.2) Designate a sub-set of paid editors "commercial editors"

This would apply to a group of paid editors who create content they aren't linked to (as opposed to, for example, employees writing about their company). This could apply either to freelancers or organisations like WikiPR. Additional restrictions could then be applied to this subset in additional proposals (without applying to all Paid Editors). Any proposals below that target this subset would be dependent on this proposal passing. Restrictions or limitations that are generalist could be !voted on specifically for "just for commercial editors".

To the degree that it would be applicable, this status would not apply to Wikimedians in Residence/WMF team members

Discussion of the idea

I am not sure either what this is supposed to mean we do or what it is intended to accomplish. Seraphimblade ^{Talk to me} 15:36, 21 July 2020 (UTC)

I've tried to clarify this above. Of itself, it isn't designed to do much, but it's preferable to create a category of editors then apply other proposals to it that actually set extra limits, rather than risk having different proposals applying to slightly different groups. Nosebagbear (talk) 16:03, 21 July 2020 (UTC)

• This is so convoluted and silly. Our paid editing policies need to be strengthened in consequence, not made more complicated to understand. Praxidicae (talk) 16:57, 21 July 2020 (UTC)
• To my knowledge, I'm the one who started using this term and encouraging people to use it (I've been saying it for years, glad it's catching on.) I don't really think we need to redesignate anything. The TOU and policies were fairly obviously not written with WiR and WMF staff in mind. All proposals about paid editing are also not written with them in mind, but it is used as a strawman to defeat any proposal about getting tough on paid editing.
  The solution to this isn't to redesignate a new subgroup with different definitions, etc. It's just to use the phrase instead of "paid" as it makes it clearer what is being talked about and avoids proposals being derailed over the spectre of an anti-paid editing zealot admin blocking the local librarian-- something that was never going to happen anyway, but gets brought up in these RfCs.
  Basically, we can just start using the term in our discussions and future policies. No need to define it further, and no need to create a new sub-group. The great thing about English is that it's usually pretty straightforward. TonyBallioni (talk) 17:03, 21 July 2020 (UTC)

  The thing is that the Wikimedia terms of use do include Wikipedians in residence under the category of paid editor, as can be seen from its frequently asked question on the subject. I don't think it's a big deal to define "commercial editor" as a paid editor minus specific exceptions. isaacl (talk) 18:21, 21 July 2020 (UTC)

  In some circumstances, yes, but usually, no: Disclosure is only necessary where compensation has been promised or received in exchange for a particular contribution. Generally being compensated as a Wikimedian in Residence is not covered by the disclosure requirement, unless there are specific edits being paid for. That said, I wouldn't oppose a clear one-line definition of a stricter policy applying to these groups. TonyBallioni (talk) 23:12, 21 July 2020 (UTC)

  Yes, quite logically, if they're being paid to make an edit, they're categorized as a paid editor. The point being that the terms of use did indeed take this scenario into consideration. isaacl (talk) 23:37, 21 July 2020 (UTC)

Restrictions/limitations

2.1) Require paid editing status to be included in any username

This would require (paid) or (paid editor) to be included at the end of any username that is a disclosed paid editor. This would enable users to identify their edits more easily. It would require significant numbers of renaming requests and may have undesired blowback at the disclosing editors.

This status does not apply to Wikimedians in Residence/WMF team members

Discussion of the idea

This could be useful. Currently, even if a paid editor is disclosed, it is necessary to visit their user page to see that. Seraphimblade ^{Talk to me} 18:55, 21 July 2020 (UTC)

I doubt this is ultimately workable. A fair chunk of paid editors don't look at the username policy in the first place and then get blocked for a promo/shared username, and users who're trying to be surreptitious aren't going to add it. —A little blue Bori v^_^v ^{Hasteur Hasteur Ha-- oh....} 20:49, 4 August 2020 (UTC)

2.2) Prevent paid editors from submitting articles to AfC

This would leave paid editors with their ability to submit edit requests on pre-existing articles (and remove vandalism etc) but prevent completely new articles being submitted through any route.

This status does not apply to Wikimedians in Residence/WMF team members

Discussion of the idea

I don't see this working. This will just lead to even more of them trying to slip new articles in as UPE. I'd much rather see them disclose and submit through AfC. Seraphimblade ^{Talk to me} 18:55, 21 July 2020 (UTC)

This is also my position. It also seems to defy a key purpose of AfC Nosebagbear (talk) 14:10, 15 October 2020 (UTC)

2.3) Prohibit paid editors from participating in policy change discussions

This would prevent paid editors from contributing to amend (or prevent amendment) of en-wiki policies/guidelines. It is not a general "wp-namespace ban". Case by case exemptions could be permitted, either for general participation or for non-!vote comment(s), where their participation is viewed as key.

This status does not apply to Wikimedians in Residence, and WMF team members so far as they can usually participate in policy discussions.

Discussion of the idea

Is this a major problem? I haven't generally seen that happening. Seraphimblade ^{Talk to me} 18:55, 21 July 2020 (UTC)

An unusually clear instance of it is happening right now. WP:VPP#Stop InternetArchiveBot from linking books→WP:BOTN#VPPOL discussion closed: linking by InternetArchiveBot. —Cryptic 12:11, 22 July 2020 (UTC)

but in what way is the contribution there by the coi editors a problem? DGG ( talk ) 00:46, 16 August 2020 (UTC)

A more relevant example, some paid editors recently !voted to delete {{Undisclosed paid}}. MarioGom (talk) 14:22, 20 May 2021 (UTC)

• This seems obvious to me, if applied to commercial paid editors. Also prohibited from taking on other governance roles (in mediating or resolving behaviors, or guiding movement affiliates) The paid editors are good at playing Nomic, and they should be kept out of rules-lawyering that directly affects them, as well as other areas of governance -- Nomic-style regulatory capture is one of our primary failure modes that would be very hard to resolve. (It has already been hard to resolve on wikis where a plurality of admins belonged to some faction that was opposed to some of the principles of free knowledge) It equally seems obvious to me that this should not apply to WiRs and the like, including your example Cryptic; especially in any cases where the organziations involved could easily be affiliates or sibling projects themselves (but for the preexistence of a separate charity or library doing archival or free knowledge work). ~~~~

2.4) Prohibit paid editors from participating in AfDs

This would prevent paid editors from contributing to Articles for Deletion (for articles they have a PAIDCOI towards).

This status does not apply to Wikimedians in Residence/WMF team members

[Discussion on whether a built-in exception where the editor is the creator of the article should be had.]

Discussion of the idea

As an alternative, I could see making disclosure of paid status in the AfD argument mandatory. If the paid editor can actually provide useful information, that's fine, and otherwise the closer can always appropriately weight comments. Seraphimblade ^{Talk to me} 18:55, 21 July 2020 (UTC)

I support it, absolutely as Afd is rife with it. How would you enforce it though? They are already UPE, so how would discover who they really are? scope_creep^Talk 19:06, 21 July 2020 (UTC)

So we get disclosed paid editors acting in AfDs as well (I realise the page is called UPE proposals, and should probably be more like "Paid Editing Proposals"), so that's one side. UPE concerns could be raised in the same way as SPA concerns. Obviously non-detection will remain an ongoing issue - that will be the case across all these proposals. I don't think AfD restrictions would cause a shift to more UPE (something which is not the case for too-aggressive restrictions), so in that sense it seems a net positive Nosebagbear (talk) 10:42, 22 July 2020 (UTC)

I'm not sure this would help. AfD isn't a vote anyway. Meritless arguments by SPAs will (should?) be given little weight as it is. ProcrastinatingReader (talk) 12:01, 29 July 2020 (UTC)

A paid editor could make a legitimate !vote, and I'd still consider it dubious. Even if they weren't specifically hired to participate in the AfD (which would be an outright canvassing breach), I'd still consider it to be a negative (e.g. if you had an employee who just generally "looked after" the article, I still think counting it as an AfD !vote would be unwise. Probably 20% of known or suspected paid editor posts in AfDs I see are somewhat policy backed, and probably 10% are reasonable ones (the others distinctly read as cribbed (!)votes) Nosebagbear (talk) 14:49, 29 July 2020 (UTC)

As a general rule, paid/COI editors have a habit of not making policy-relevant arguments in deletion debates, usually defaulting to one of the many "loser" arguments (primarily "I like it" or "Other stuff exists"). Any AfD closer worth their salt will discount their arguments. This is a solution seeking a problem. —A little blue Bori v^_^v ^{Hasteur Hasteur Ha-- oh....} 20:56, 4 August 2020 (UTC)

• I think paid editors should be welcome to bring up sources to AFD. The main ongoing issue with AFD and paid editing comes from UPE operations with many sockpuppets, so I'm not sure this particular proposal would have any positive impact. However, I would support a proposal for paid editors to be required to disclose on AFD when they participate, in addition to their disclosures elsewhere. MarioGom (talk) 14:26, 20 May 2021 (UTC)

2.5) Dedicated accounts and userrights

Established editors taking up paid editing must create a legitimate alternate account and link the two accounts and clarify the difference. Paid editing must occur on the alternate account. Non-paid editing, so far as is viable, should occur on the primary account. The alternate account may immediately receive confirmed/extended-confirmed user-rights (if the original account holds them) but may not hold any other user-rights.

This status does not apply to Wikimedians in Residence/WMF team members

Discussion of the idea

• It wouldn't work as there would be a fairly heavy and cumbersome administrative load and we already have insufficient admin numbers to enforce. How would you know they were using correct account and how would it be enforced? scope_creep^Talk 18:38, 21 July 2020 (UTC)
• I could see this potentially being workable. Seraphimblade ^{Talk to me} 18:56, 21 July 2020 (UTC)
• "Does not apply to WMF members" actually contradicts current policy, doesn't it? I think WMF people all have alt accounts for their volunteer activity. - Bri.public (talk) 16:58, 27 July 2020 (UTC)

  @Bri.public: - I'd actually wondered that, I didn't want to find there were circumstances where non-team accounts were legitimately used. If however we're confident that it's the case then obviously it can just be removed. Nosebagbear (talk) 17:55, 27 July 2020 (UTC)

  There is Category:Wikimedia Foundation staff but I don't find a matching policy that they must use the (WMF) tag. So I'm not sure whether it is a policy or just a convention. Maybe we should ask a community-facing WMF office. Bri.public (talk) 18:16, 27 July 2020 (UTC)

  Bri, in the wake of the Superprotect debacle, where there was some confusion over who was WMF or whether someone was speaking in their official or volunteer capacity, the WMF agreed that going forward all staff acting or speaking in their official capacity would use an account marked with "(WMF)" in the name. They have, to the best of my knowledge, abided by that since, although I don't know if it was made part of their official policy anywhere. Seraphimblade ^{Talk to me} 20:19, 27 July 2020 (UTC)

  Found it: "we will put in place a clear distinction between work and personal accounts for all WMF employees by September 15" – LilaTretikov (WMF) 19 August 2014 [1]. I'm surprised this isn't codified somewhere, or it's well hidden, or I'm just not very good at looking up policies. - Bri.public (talk) 22:15, 27 July 2020 (UTC)

  I think the more important part that WMF staff should be exempted from is the prohibition on holding manually-assigned user rights. I feel the community would generally prefer there be more experienced English Wikipedia editors amongst the Wikimedia Foundation staff. isaacl (talk) 22:39, 27 July 2020 (UTC)

  You're absolutely right. Nosebagbear (talk) 23:14, 27 July 2020 (UTC)

2.6) 500/30 article creation prohibition in "spammy areas"

Editors must be be extended-confirmed to create new articles in "spammy areas" which tend to see lots of paid editing. These areas being: blockchain and crypto, speculative financial instruments, medical topics, online gambling, startups, businesspeople and investors in startups, digital marketing (incl SEO), insurance/debt companies. (list taken from User:MER-C/Paid2019).

Optionally: editors not meeting this criteria could be allowed to use AfC. Or the prohibition could be strict and articles could be deleted. Up to the discretion of an admin (similar to ArbCom's Israel-Palestine sanctions, I believe). ProcrastinatingReader (talk) 13:52, 30 August 2020 (UTC)

Comments on the proposal structure/summary

• So I'd be very mildly in favour of this in some areas - I think the list above is a bit too broad, but could back crypto and so on, that is, areas where genuinely most of the creations are spammy (so not medical articles). I'd be firmly against prohibiting absolutely - AfC must remain an option. Nosebagbear (talk) 13:55, 30 August 2020 (UTC)

  Striking medical topics. I was thinking more along the lines of psuedoscience, cosmetic surgery, etc., but I now see the typical medical topics. In any case, better to start with a simpler list. Any issue with the remaining topics, Nosebagbear? ProcrastinatingReader (talk) 16:33, 30 August 2020 (UTC)

  Hmm, startups might want to be defined as "less than 1 year" to stop it becoming overbroad, even if others leak in. I'm kind of neutral on the insurance/debt companies - I'm interested in what others have to say on that one. Nosebagbear (talk) 16:38, 30 August 2020 (UTC)

Enforcement

3.1) Creation of a new CSD that allows deletion of articles & drafts that were created by an undisclosed paid editor

Currently only creations by an editor who has already been blocked can be deleted. This would allow deletion of content created by an UPE that was made before they were identified as such. It could not be used if another, legitimate, editor had made any substantive contribution. Content deleted by this route would be liable to REFUNDing for any editor in good standing.

A degree of practical good faith in cases where a creator is not a disclosed paid editor but does not give any indication of attempting to conceal their status (e.g. their username is their company name) should be applied, seeking disclosure rather than immediate nomination.

The aim here is to reduce the motivation for paid editors (particularly "commercial editors") from being able to achieve their goals even if they are ultimately blocked. Against this is the loss of potentially notable content that otherwise meets the rules other than for the status of its creator. In effect, the proposal is a "greater good" dispute.

Discussion of the idea

Redundant. Undisclosed paid editing is already banned, so known UPE is already covered by G5. The trick is actually proving it, especially to do so without violating the outing rules. (I would like to see a proposal here regarding whether revealing evidence that someone is paid does indeed constitute outing, or whether rules in that regard should be relaxed for showing evidence that a denial of being paid is false.) Seraphimblade ^{Talk to me} 19:02, 21 July 2020 (UTC)

@Seraphimblade: I've seen numerous cases where an editor has carried out various UPE, then been caught, and been blocked. However G5 only applies to edits made after an editor has been blocked, not merely after they did something that warranted blocking. So known creations by a (now known) UPE are not deleted as G5 doesn't apply. Nosebagbear (talk) 21:38, 21 July 2020 (UTC)

That's flat incorrect. G5 includes breaches of limited or partial bans, such as a topic banned editor creating an article in the area from which they're topic banned. In this case, every editor is subject to a "topic ban" on UPE, so G5 absolutely does apply to it. It's editing in defiance of a ban. Seraphimblade ^{Talk to me} 23:35, 21 July 2020 (UTC)

Community interpretation seems to disagree with that - otherwise we wouldn't have multiple prior discussions on this proposal. If the community can be convinced this is the interpretation that's fine, but it definitely isn't the default judgement. Most people seem to interpret the "topic ban" as a rule being broken, rather than a standing TBAN, as it were. Nosebagbear (talk) 23:44, 21 July 2020 (UTC)

I have reviewed what G5 currently actually says, and it says nothing about excluding UPE. (That's in contrast to something like, say, A7, where specific exclusions such as schools are made from it.) If it's desired to exclude UPE from G5, someone can make a proposal to do that, but until and unless they successfully do, UPE is covered under G5. Seraphimblade ^{Talk to me} 23:55, 21 July 2020 (UTC)

UPE is covered under G5. I wish it was. In practice, it is not. There are several declined G5 on articles created by UPE based on the block date vs article creation date. MarioGom (talk) 14:31, 20 May 2021 (UTC)

It's worth making this explicit. If we're limited to indef-blocking three-edit accounts that would've been thrown away anyway, anything else we come up with here is completely toothless. —Cryptic 12:14, 22 July 2020 (UTC)

• Agreed that it should be redundant, and that it should be explicit. ~~~~

3.2) Editors found to have committed undisclosed paid editing may have all prior substantive edits removed

Unlike proposal 3.1, which focuses purely on created articles/drafts, this would allow any substantive edit be removed. This both allows more and less content to be removed.

Editors may make their own judgement calls on removing/retaining beneficial content, and the editor's content may be removed from any article/draft, except where that would negatively affect another editor's contributions.

The aim here is to reduce the motivation for paid editors (particularly "commercial editors") from being able to achieve their goals even if they are ultimately blocked. Against this is the loss of potentially notable content that otherwise meets the rules other than for the status of its creator. In effect, the proposal is a "greater good" dispute.

It prevents paid editors who avoid writing content from scratch being in a loophole that 3.1 would create, however, the possibility for problematic removals is higher when considering pages with edits from other users.

Discussion of the idea

Exactly this, and that's why we have G5 in general. Banned edits can be reverted, and banned creations deleted, regardless of their quality, to deter banned editors from evading the ban. This would be the exact same rationale—get caught in UPE, then it all goes up in smoke and now your client is calling you and screaming for a refund. Seraphimblade ^{Talk to me} 19:02, 21 July 2020 (UTC)

• Absolutely support this, maybe as addenda to WP:REVERTBAN. This does not mean that all edits should be reverted automatically, but any reviewing editor should be able to err on the side of revert. See User:Blablubbs/Wolfram and User:MarioGom/LoboReview for some large review examples. MarioGom (talk) 10:56, 9 September 2021 (UTC)

3.3) Authorisation of "mystery shoppers" to counter UPE offerings

French Wikipedia recently used several volunteers who replied to a significant number of paid editing offers, seeking additional information (article examples, demonstration of user-rights, confirmation of username control etc). This allows problematic accounts (including, where appropriate, CU-involvement), especially those with userrights, and articles to be identified.

Significant further discussion on permissible methodology, selection of willing volunteers, and oversight (most likely through ARBCOM) would need to be had where this proposal to get an initial indication of community support.

Discussion of the idea

This has already been happening, but on an ad hoc basis. I think developing rules and a system around it is a very wise suggestion. Seraphimblade ^{Talk to me} 19:02, 21 July 2020 (UTC)

Yes, given that I've been tempted to act on it before, I'd be shocked if no-one has. However, community authorisation and some rules to mitigate issues would be major pluses. If even 3% came with a risk of being blocked for reaching out that would force a significant reticence in off-wiki promotionalism. Nosebagbear (talk) 21:42, 21 July 2020 (UTC)

@Seraphimblade and Nosebagbear: It's quite important that we do this - in a separate policy or at WP:Harass. The reason that investigating open websites which say "we edit Wikipedia for you" is not done is that people think that reporting it is against our rules, e.g. harassment, outing. This could be a very simple policy. For example "Editors may investigate firms and their websites that advertise Wikipedia editing services, to ensure that they are following our policies, e.g on WP:Paid editing disclosure." Without something written in policy we're tying both hands behind our backs. I'll guarantee that The Signpost would nave a carefully written article on these websites, if this was written into policy, but otherwise we are very limited in what we can say - the current lack of policy has an incredible chilling effect. We have reported on these websites, but only in very clear-cut situations, e.g. when the French editors did the investigation and reported it on FR:wiki. Since this was already on Wiki and posted by admins(!) I felt there was no problem. Similarly, in the current issue of WP:POST, I didn't have a problem reporting that a website used Jimbo's name and photo in an advertisement - because it had already been reported on Jimbo's talk page by an admin. Even then there were some special precautions I took or things I felt I had to omit. More important thanThe Signpost's difficulties, of course, is difficulties in enforcement when people aren't sure if they can report this type of things. Smallbones_(smalltalk) 12:39, 4 August 2020 (UTC)

I wanted to give some examples of confusing policy that inhibits *any investigation* of these sites. Most of it is at WP:OUTING which has very severe penalties like (approx) "any admin may permanently block you immediately at any time for messing up on this". The whole of the outing main section is like this, reading to me like, *mentioning any editor's "legal name, ... workplace address, job title and work organisation, telephone number, ... other contact information, or photograph, whether such information is accurate or not," unless they've posted it and not removed it on EN.wiki. There are exemptions to this, example, fine point, if and or but, contradiction, outside websites, and everybody else is covered, exception, fine point, contradiction." etc, etc.

Other sections below WP:OUTING are just as bad, e.g. WP:EMAILPOST, WP:OWH, WP:HNE. Nobody wants to disable protections against harassment, but I'd certainly like an understandable exposition of what those protections are. And especially I'd like to have something like, "websites of paid editing firms may be investigated and reported on-Wiki, as well as other websites where a company appears to be saying "we regularly violate Wikipedia's rules." Otherwise investigating paid editing websites and actually telling anybody about it is a mine field just waiting to blow your leg off. Smallbones_(smalltalk) 14:30, 4 August 2020 (UTC)

I agree. It's got enough issues that I'd add a few of the common risk areas (Outing, most notably) that currently discourage this and see if there's a general consensus to enable it if backed generally then go to a proper drafting stage, that will need a fair amount of different groups weighing in. Nosebagbear (talk) 20:06, 4 August 2020 (UTC)

• I am not sure the direction this may go but with current policies mentioned above I would not even bother with any "investigation" because I like both my Wikipedia legs. It would seem that if there was a link to a site that in any way states "We will edit Wikipedia for you", there would be some way to investigate this without fear. I am not sure if WP:Harass would be it. -- Otr500 (talk) 14:56, 23 August 2020 (UTC)
• The problem with this is that we have to ask people to be dishonest in order to find paid editors. They need to lie, falsely represent their interest, and attempt to gain information about paid editors through those false claims. Yet as a project we rely on honsty - we ask that people be honest and upfront about COIs; not to use multiple accounts even though it would be very easy to get away with it to win disputes; honestly and accuratly represent sources; openly declare when they are being paid; and be honest about their qualifications. I don't feel comfortable with the ethics of insisting on honesty from contributors while the same time designing something that requries dishonesty to operate. - Bilby (talk) 10:27, 30 August 2020 (UTC)

You could do a reasonable job of mystery shopping without actually lying, though it would obviously require omission of the truth. Just showing interest and asking for examples and so on. I absolutely get your concerns, and yes, I suspect lying could well be needed to get the more cautious examples - I even have some support for that viewpoint. Hmm, interesting consideration. Nosebagbear (talk) 13:57, 30 August 2020 (UTC)

• Full support. This will have a far greater impact than any policy change. Let's please make this happen. {{u|Sdkb}} ^talk 20:48, 15 October 2020 (UTC)
• Yes, this seems a high priority change. ~~~~

3.4) Persistent UPE operations should be escalated to the Wikimedia Foundation

There is already the precedent of the WMF sending a cease and desist letter to Status Labs. Some large UPE operations are organized from the United States by (theoretically) reputable companies. Some of these might be deterred by cease and desist letters or other action by the WMF. This will obviously not work for UPE operations from random companies in remote or hostile jurisdictions, but could be feasible with American companies.

The exact form of this proposal would need direct collaboration from the WMF.

Comments on the proposal structure/summary

• I think the highest escalation should be to a community group, for robustness and governance and scale. But there could be templated responses that WMF and other regional entities are ready to send in support, where that is likely to be helpful. – SJ + 22:58, 13 September 2021 (UTC)