Wikipedia:Reliable sources/Noticeboard/Archive 348

Voting for question 1 (WikiLeaks)

• Option 2. Avoid using WikiLeaks for ongoing controversies or if there is coverage by RS. All for using WikiLeaks as the only source, particularly when English courts accept them as evidence, if verifiable information (facts) are mentioned in the document; opinions should be evaluated for being WP:DUE. Their selection may exhibit owners' bias, but taken one-by-one, the documents seem to be all right, and no one has shown that any of the documents were forged or doctored, as Alaexis correctly points out. Szmenderowiecki (talk) 04:53, 12 May 2021 (UTC)
• Option 2. I don't think WikiLeaks ought to be cited directly in most circumstances it tends to cover (usually if it's notable enough RS picks up the slack) but for small clerical bits and bobs of foreign policy I don't see an issue. Paragon Deku (talk) 05:16, 12 May 2021 (UTC)
• Option 4. Wikileaks, itself, performs no fact-checking or verification. Therefore, things there do not pass the definition of "published" in WP:RS or WP:OR and cannot be cited directly under any circumstances, fullstop. I would consider Wikileaks (when used alone) a remove-on-sight source and I'm baffled that anyone would argue otherwise - it is no different from self-publishing in this context. If a secondary source covers it, we can rely on what they say, but only for the parts they specifically mention, since only those parts have been published; the argument, which some people are trying to make in the linked thread, that we could say "well, this trove of documents is validated in this source, therefore we can go through it and pull out anything we please even if it has no other coverage" is straightforward WP:OR. --Aquillion (talk) 06:17, 12 May 2021 (UTC)

WikiLeaks does, in fact, verify the authenticity of documents before it publishes them. The most famous leaks published by WikiLeaks are all widely acknowledged to be genuine: the Camp Delta Standard Operating Procedures, the US diplomatic cables, the Afghan War Diary, the Iraq War Logs, the Guantánamo Files and the Syria Files. Several of these publications were carried out in collaboration with major international newspapers, including Le Monde, The Guardian, Der Spiegel, the New York Times and El País. I don't believe any document published by WikiLeaks has ever been shown to be fake. If you have seen evidence to the contrary, then please post it. Otherwise, it looks like WikiLeaks has a very strong track record of authenticating documents before publication. Whether those documents are usable is a completely different matter, because they may be primary sources, they may express opinions, etc. But they are genuine documents. -Thucydides411 (talk) 11:02, 13 May 2021 (UTC)

• Option 2. Generally reliable per WP:USEBYOTHERS and no evidence of tampering. Probably should not be used as the only source for controversial statements or in BLP context and in general should be treated as a primary source. Alaexis_¿question? 06:19, 12 May 2021 (UTC)
• Option 4. I completely agree with Aquillion's rationale above. ElKevbo (talk) 06:44, 12 May 2021 (UTC)
• Option 2. I agree it fits WP:USEBYOTHERS. There are some cases when it is used to supplement or reinforce claims made elsewhere in investigative journalism and whatnot. In that case, you should probably refer to the sources doing that though I suppose it might not be necessary to also link to Wikileaks in that case. FelipeFritschF (talk) 07:21, 12 May 2021 (UTC)
• Option 4. Wikileaks is an indiscriminate collection of communications and internal files leaked by a website with a very specific political agenda. The documents themselves are not official in any sense of the word: they have only been drafted by government employees, often with very little oversight and—obviously—no peer-review or editorial standards. Moreover, they have been covered by many, many press articles from highly-reliable source: if editors cannot find a press article covering the leak in question, this should be an indication that it is dubious. The WP:USEBYOTHERS argument is not applicable here, because Wikileaks is, functionally, documentation center: it would be like citing files from historic archives, directly, on the grounds that professional historians use them. JBchrch (talk) 10:15, 12 May 2021 (UTC)

  Unless I am misunderstanding your use of the phrase "historic archives", such files are used quite frequently as references in all sorts of articles about old stuff. jp×g 20:04, 12 May 2021 (UTC)

  I am mostly addressing the USEDBYOTHERS argument made above. The fact that experienced journalists are using leaked emails as the basis of their reporting does not make it acceptable (in my view) for editors to use such materials as sources on the basis of USEDBYOTHERS. On your point about archive documents: yes, you sometimes see them, but (as I understand) the real standard (i.e. the one used at WP:FA) is that it's not the recommended way to source articles. JBchrch (talk) 00:01, 13 May 2021 (UTC)

  Never trust anything you read from unreliable sources, such as Wikipedia. According to the Wikipedia article you are citing, Wikileaks has promoted a conspiracy theory "that Hillary Clinton wanted to drone strike Assange". On its face, it looks like this content was added by some POV pusher who did not consider the context.

  If one actually reads the source, it is a Snopes fact-check about a Tweet by Wikileaks. The tweet actually promotes a claim in a report by True Pundit, which attributes the droning claim to "State Department sources". Snopes considered the claim questionable, but was unable to disprove it, rating the claim "Unproven". Clinton did not categorically deny the claim; according to Snopes, Clinton did not "recall any joke ... [reference to targeting Assange with a drone] would have been a joke". Moreover, Snopes based its analysis on some governmental documents published by Wikileaks without questioning their authenticity, which undercuts your argument that Wikileaks is unreliable. Politrukki (talk) 12:30, 31 May 2021 (UTC)

• Option 2. We should treat Wikileaks closer to how we treat a publisher of user-generated content (e.g. YouTube) than how we treat a publisher of in-house journalistic works (e.g. a Newspaper). Content on Wikileaks is a mix of verified and unverified, notable and non-notable, works by a massive range of authors some of whom are subject-matter experts, some of whom are random people on the internet. In most cases the copy of Wikileaks can be regarded as an accurate copy of the primary source documents, without guarantee (in most cases) that every document that is part of a set is present, but the reliability of the documents themselves must be evaluated individually. Thryduulf (talk) 10:19, 12 May 2021 (UTC)
• Option 2 I agree with everything Thryduulf says above. It would be taken on a case by case basis and attributed appropriately. Spudlace (talk) 10:40, 12 May 2021 (UTC)
• Option 2 Should ONLY be used in conjunction with reliable secondary sources that have vetted the specific information being cited. Basically, if someone like the New York Times has written an article about something in Wikileaks, then it may be OK to also cite Wikileaks alongside it to cite a specific quote or paraphrase, HOWEVER, it should be treated like a primary source otherwise, and should also never be used to cite something that has not already been vetted in reliable sources which are also cited in Wikipedia. I am very leery of using results of random data scrapes from Wikileaks and accepting the results of that as sufficient to cite some statement at Wikipedia, no matter how banal. If it only exists in Wikileaks, and no other reliable source has vetted it, it's a hard no from me. --Jayron32 12:36, 12 May 2021 (UTC)
• Option 3 or 4 In principle, I think there might be cases where a document posted to Wikileaks is useable, but in practice, such situations are very difficult, perhaps even impossible to find or even formulate as hypotheticals.

With respect to the claim about documents on WL being primary: in many cases, they're quite clearly not primary. A recent example I saw was a cable purporting to be from a US embassy describing the membership of a Laotian political committee. It's clearly not a primary source with regards to that, as it doesn't purport to be from the Laotian government, nor any member of it, but from a US embassy; undoubtedly a third party.

However, the reliability of documents on WL is highly debatable. There's no system of checks and balances, no chain of custody, and usually no way for a WP editor to verify the accuracy or provenance of the documents. They might and indeed probably are what they purport to be, but we have nothing but WL's rabid anti-secrecy stance to evince that. But we also know that WL has a right-wing, or at least conspiratorial bias, and numerous connections to Russian anti-democratic cyberwarfare actors. We even know that they don't always support their own principles, as WL and Assange were notoriously critical of the Panama Papers. We also know that they claimed that the Clinton email leak did not come from a Russian source, when virtually every cybersecurity expert out there was in agreement that it did.

Even though I generally believe that the documents on WL are what they purport to believe, I cannot dismiss the possibility that WL would allow or even engage in the forging of leaked documents, and they provide no mechanism to assure us that they haven't.

So in any case in which a document leaked to WL is to be cited, I would instead seek to cite coverage of that document in reliable sources, instead. At the very least, I would cite both the document on WL, and the RS that vetted the document. If no RS has vetted any particular document, then I would not cite it at all, absent a compelling (and hitherto unimagined by me) argument. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 16:16, 12 May 2021 (UTC)

• Option 1 (WikiLeaks has a strong record of validating documents) and Option 2 (additional considerations apply when using these documents). Option 2 is the obvious answer. WikiLeaks hosts various types of leaked documents. It's impossible to give one single rating to all the documents, because they're so different from one another. If ever there was a case of "additional considerations apply", this is it. WikiLeaks has a very strong record of verifying the authenticity of the documents it publishes, and I don't think there is any known case of WikiLeaks having published fake documents (contrary to the evidence-free speculation by some editors above). Some of WikiLeaks' publications are extremely well known and have been vetted by numerous other organizations: the US diplomatic cables, the Afghan War Diary, the Iraq War Logs and the Syria Files come to mind. The reliability of the claims made in any of these documents would have be be determined on a case-by-case basis, taking into account who wrote the document, the claim being sourced, etc. Most of the documents are also primary sources, which would obviously affect how they can be used. As I said, additional considerations apply. -Thucydides411 (talk) 17:24, 12 May 2021 (UTC) (Updated based on Szmenderowiecki's clarification of what the options mean. -Thucydides411 (talk) 15:34, 13 May 2021 (UTC))
• Option 4, per Aquillion. --JBL (talk) 17:57, 12 May 2021 (UTC)
• Option 2. I don't see any evidence that WikiLeaks has falsified information. I see the claim that they are "rabid", but not that they are unreliable; the political opinions of the people who run a website do not magically make the content on it unusable. Sure, it is great to back it up with a second reference to another RS, but that's true of basically anything. The claim that "well, I think the New York Times is trash because they're a bunch of libs" doesn't cast substantial doubt on the fact that, generally, the NYT is a reliable source for factual statements; I don't see why it is any different for WL. jp×g 20:04, 12 May 2021 (UTC)
• Option 2 (with overlap in opinion on the question to Thucydides411). And I'm just spitballing here, but — I would recommend looking at potentially some sort of time divider similar to Newsweek here, because I think they had a much better reputation for integrity pre-2016 (or so). --Chillabit (talk) 20:37, 12 May 2021 (UTC)
• Option 4. I have had enough of this bullshit. This is not actually a RS quesiton, it's an attempt by the Assangites to crowbar Wikileaks into the project in defiance of a blindingly obvious WP:UNDUE failure, but their determination makes it necessary to be unambiguous. No. We absolutely do not include stolen copies of primary sources published on a site that has been a Russian intelligence asset for at least five years, because of course we fucking don't. Guy (help! - typo?) 21:19, 12 May 2021 (UTC)
• Option 1. Wikileaks has a process for verifying documents prior to publication.^[1] It has published an enormous number of documents and, while there have been general claims that Wikileaks has published fake documents, I haven't seen a case where a specific fake document was identified. It is clear why some people or organisations would like to claim Wikileaks is unreliable. The documents themselves should be treated as primary sources. Any statements or claims made within the documents published by Wikileaks may be erroneous but that is a separate matter. Burrobert (talk) 22:27, 12 May 2021 (UTC)
• Option 1 per Burrobert Tyrone Madera (talk) 23:45, 2 June 2021 (UTC)

• Burrobert, sure, we know: as far as you are concerned, being on Wikileaks means that something is true, and of unambiguous significance. Only issue is, one in eight of your 4,000 edits is to Julian Assange (edit | talk | history | protect | delete | links | watch | logs | views), or Talk, and 100% of them are Assangite boosterism. Guy (help! - typo?) 22:59, 12 May 2021 (UTC)

"being on Wikileaks means that something is true, and of unambiguous significance": I said that a document being on Wikileaks means we can trust that it is genuine. The claims made within the document are a separate issue. I didn't comment on the significance of any document and the RfC is not asking us to address that issue. The significance of any particular document on Wikileaks should be determined in accordance with existing procedures for treating primary documents. Burrobert (talk) 23:09, 12 May 2021 (UTC)

The source cited here for its fact-checking process (which is from 2009 and so possibly out of date for more recent material anyway) does not inspire confidence: The number of people involved in the verification process, as with the rest of Wikileaks, is unclear. But Wikileaks claims to have published 1.2 million documents in three years. This means its - presumably extensive - team of volunteers receives, verifies and publishes over 1,000 documents every day... There is fake content on Wikileaks. A whistleblower, who asked to remain anonymous, admitted to submitting fabricated documents to Wikileaks to see what it would do. The documents were flagged as potential fakes, but the whistleblower felt that the decision to publish the documents had "an impact on their credibility"... most of the members of the advisory board to whom Wired spoke admitted that they had little involvement with Wikileaks, and have not done much "advising". "I'm not really sure what the advisory board means," says Ben Laurie, a computer- security expert and member of the board "since before the beginning". "It's as mysterious as the rest of Wikileaks."... Phillip Adams, an Australian journalist, is listed as an advisor. But he told Wired that he had "resigned early on because of workload and health issues". BobFromBrockley (talk) 20:30, 13 May 2021 (UTC)

An anonymous person claims that they submitted fake documents, which WikiLeaks correctly flagged as fake. Meanwhile, all of WikiLeaks' major publications are widely considered to be genuine. These include the Camp Delta Standard Operating Procedures, the US diplomatic cables, the Afghan War Diary, the Iraq War Logs, the Guantánamo Files and the Syria Files. For many of these, WikiLeaks worked with major newspapers, such as Le Monde, El País and the New York Times. Look, if you want to argue that WikiLeaks cannot be trusted to validate documents, then you'll have to address the fact that its major publications are widely considered genuine, and you'll have to provide some actual evidence that WikiLeaks is unreliable. I haven't seen anyone in this thread do so yet, which makes the "Option 4" votes quite puzzling. -Thucydides411 (talk) 20:41, 13 May 2021 (UTC)

• Option 4 Per Aquillion and others. Wikileaks has never been a reliable source, and has only gotten less reliable as Assange's infatuation with Trump grew. Security experts have repeatedly cautioned about accepting Wikileaks dumps at face value, and given Wikileaks intentional obfuscation, and outright lies, about its sources, which it weaponizes to achieve its political goals as in the Seth Rich case, it should be abundantly clear that they cannot be trusted. Any outlet that intentionally weaponizes disinformation should not even be considered as a source for Wikipedia.NonReproBlue (talk) 01:34, 13 May 2021 (UTC)
• Option 4. This is a questionable source, a disorderly collection of WP:PRIMARY claims some of which may be outright wrong of very difficult to properly interpret. In any event, one needs other secondary RS that provide proper context. But if there are such RS, then the claim can be cited with a reference to the secondary RS, not Wikileaks themselves. My very best wishes (talk) 03:45, 13 May 2021 (UTC)
• Option 2 - it is generally concluded by reliable sources that WikiLeaks is accurately portraying the documents as they are there. However, care should be taken as to which documents are used, and they are all in any case primary sources for subject matter talked about, unless they are finished copies of documents that summarize other sources - and even then, they're less usable than other secondary sources. Their obvious bias doesn't matter - we don't require sources to be unbiased at all - and in fact there are multiple obviously biased sources that are perfectly reliable sources (looking at CNN, as an obviously biased but still reliable source, as an example). Note that the "published" argument does not apply either - because "published" doesn't mean that it's accessible to the public - and completed government documents are not unreliable simply because they are or were classified and thus never published in a public source. As a primary source, documents from WikiLeaks can be used - but I echo the concerns of many here who have said that it would be preferable to find coverage of the documents and cite that instead - if only for the added encyclopedic information such coverage may provide. In a case where no other coverage exists but a document on WikiLeaks expresses a significant and encyclopedic view, it can be cited as reliable. No evidence has been provided that WikiLeaks is systematically altering documents or forging information, and in fact reliable sources don't believe they do so. TLDR: see jzg's !vote. -bɜ:ʳkənhɪmez (User/say hi!) 03:51, 13 May 2021 (UTC)
• Option 4 Per Guy & MVBW, its potentially a useful research tool, but it should at no point be cited as a source in an article. Only in death does duty end (talk) 07:28, 13 May 2021 (UTC)
• Option 2 - Seems fairly straightforward option. Basically treat it as a primary source with all the considerations that go with that. RS seem to treat them as reliable for authentic documents. Lacking any substantial reasoning beyond "Assange/Trump BAD!" I see no reason to black list them or treat them as unreliable for what they are. On the contrary, as brought up by others above, their repeated use by other RS is a strong indication that they would be acceptable. Just have to keep in mind the primary nature and careful of UNDUE. PackMecEng (talk) 11:28, 13 May 2021 (UTC)
• Option 3 They seem to (or have least been heavily accused) of just info-dumping. They may well all be true, in that they are real documents, but not that what is contained within those documents is true (after all they published the Xenu bad SF story).Slatersteven (talk) 12:45, 13 May 2021 (UTC)
• WikiLeaks is not reliable. If there's something interesting there, let journalists do their job and then cite them - First, the question is bizarre. This is formatted like a typical source RfC but the question isn't about reliability for Wikipeda, but whether it's "reliable for publication of genuine government documents". That's why I didn't just choose an option. Look, WikiLeaks is at best just a host for documents/uploads like Scribd or Etherpad or Dropbox or whatever. Add to that questions over authenticity (no, we don't need to come to a decision about whether or not they're genuine to know that there have been a lot of questions raised in reliable sources) and of course we shouldn't cite it. If it has something useful, let someone else vet it and link to it. — Rhododendrites ^talk \\ 14:33, 13 May 2021 (UTC)
• RFC choices aren't a good fit in this case. I agree with Rhododendrites above. This RfC is not a good fit since Wikileaks is not so much a publisher as a repository or primary source. A document on Wikileaks may be cited but only if a RS has discussed it. If the WSJ discusses contents of a document on Wikileaks then it may be appropriate to also link to the document. We might do something similar with a statement from the SEC. "The SEC released a statement saying they opened an investigation [cite RS, cite SEC statement on SEC page]". An edit like this is OK "According to the NYT documents released by Wikileaks showed the State Department issued a request to... [cite NYT, cite specific wikileak document <- must be clear from RS this is the correct document]" In this case the wikileaks document is a supplement to the RS's statements. It isn't a requirement. It would never be OK to cite the Wikileaks document absent a RS. For this reason I can see why editors have picked both option 2 and option 4. Springee (talk) 14:54, 13 May 2021 (UTC)
• Option 1-2, depending on what the question is, exactly. Yes, Wikileaks documents are genuine government documents (or sometimes documents from banks or other institutions). Those documents themselves may contain unreliable or inaccurate information. So in general, information on Wikipedia that's sourced to Wikileaks documents should be attributed to Wikileaks and the government document. The only case where attribution may be dropped is when the information has been verified elsewhere (e.g. by a reliable secondary or tertiary source). In that case however we may still want to attribute to the government document, particularly if its publication was the reason the information came to be more widely known. -Darouet (talk) 15:08, 13 May 2021 (UTC)
• Option 2 - First: Wikileaks is not a source, but a host site for primary documents. Wikileaks is more like publisher than an author. That said - Since the possibility exists that a document uploaded to Wikileaks may have been edited or changed from the original, we can not rely on the version hosted on Wikileaks for information. We can ONLY rely on the original, or copies that have been verified to be “true and accurate copies” of the original (example, copies that have been submitted as evidence in a court case). Now... if the version on Wikileaks HAS been compared to the original, and can be verified to be “true and accurate”, THEN we can cite the original and use the version on Wikileaks as a courtesy link.

There is one exception to this. IF a document appearing on Wikileaks is itself the subject of discussion by independent sources (say in a news story about was leaked), the version on Wikileaks can be cited as a primary source for itself (ie the text that appears on Wikileaks). The key is that it must NOT be cited as if it were the original document. Instead, it should be cited as a separate document on its own - with appropriate attribution (example: “Text of document downloaded to Wikileaks, purporting to be document XYZ” not “Text of document XYZ”). Blueboar (talk) 15:52, 13 May 2021 (UTC)

Sure, if a Wikileaks document was cited and discussed in other sources, then the claim can be cited with a reference to other sources (+ the courtesy link), but I do not see this as "option 2" when the source (Wikileaks) is regarded as an RS by itself. My very best wishes (talk) 16:22, 15 May 2021 (UTC)

• Option 4 seems to be the best fit out of the possibilities given, but I'll concur with the sentiments above that it's kind of a weird question to ask. Like Rhododendrites said, it's like holding an RfC for the reliability of Scribd. Let the journalists do their job, after which we can do ours. XOR'easter (talk) 18:38, 13 May 2021 (UTC)
• Option 4 - WikiLeaks would seem to be the quintessential case of an unreliable source. The documents it hosts are admittedly stolen, and we have no confirmed information as to what, if any, steps have been taken to ensure that they are authentically sourced and unaltered. Indeed, WikiLeaks itself accepts anonymous submissions. The face of WikiLeaks is Julian Assange, who has been on the run from the law for years. The mere fact that, in some cases, reliable sources have used particular WikiLeaks documents that they believe they have been able to authenticate seems like weak justification indeed for treating WikiLeaks as reliable. If there are particular documents that it is appropriate to cite, they should be cited to the reliable sources discussing them, not to WikiLeaks. (In such a case, I would not object to a link to the document discussed, but the document itself should not be cited for anything not in the reliable source.) John M Baker (talk) 20:09, 13 May 2021 (UTC)
• Leaning to option 3 or 4. Not totally sure yet which option is best, but two comments: 1) WP:USEBYOTHERS does not apply, for the same reason that we would not say that Donald Trump is a reliable source because he has been quoted by reliable news sources. Wikileaks material has certainly often been reported on, but (with the exception of collaborations where e.g. NYT and Guardian were able to themselves verify particular dumps) the reporting typically adds caveats. 2) This is probably too obvious to be worth saying, but the editorial material by Wikileaks itself 9as opposed to leaked material in their archives) should be treated with particular caution. For instance, its recent dumps of small batchs of highly redacted and by themselves confusing OPCW documents about the Douma chemical attack was accompanied by long editorial introductions explaining them which actually contained several errors as well as a very slanted interpretation. BobFromBrockley (talk) 20:47, 13 May 2021 (UTC)

To point 2): question (1) was not asking about the editorials that accompany the documents, only the documents themselves. What Assange says they might mean and the conclusions he says we can draw, given his strong political views, is WP:UNDUE, or, for some tastes, even WP:FRINGE. You can reformulate question 1 thus: Can we trust what WikiLeaks says are government documents to be genuine? As to point 1), I address it in the discussion. Szmenderowiecki (talk) 00:26, 14 May 2021 (UTC)

Szmenderowiecki It's cler that the question as posed here was about the documents, but when this RfC is closed, if Wikileaks is deemed reliable in any way it will be vital for the closing statement to be very clear that that decision refers to the documents in the archive and that what Wikileaks says in its own voice should not necessarily therefore be deemed reliable but treated as opinion generally is. BobFromBrockley (talk) 14:15, 14 May 2021 (UTC)

Fair enough. I just wanted to make sure that you separate the documents themselves from their interpretations by Assange/WikiLeaks staff while making summary judgment on reliability of the website (which you do), and also to warn other commenters that we shouldn't conflate these. Have a good day. Szmenderowiecki (talk) 14:38, 14 May 2021 (UTC)

• Option 4 Wikileaks is essentially Project Veritas on a global scale...private individuals with no methods of fact-checking, accuracy, or verification, claiming to be The Ones Who Show You The Truth. IF actual reliable sourced have vetted a piece of info originating at Wikileaks and voice for its accuracy, then it is fine. But at that point, the question of citing Wikileaks itself is moot. Zaathras (talk) 21:31, 13 May 2021 (UTC)

  Bellingcat can also be characterised as private individuals who claim to know the truth. Now they are considered reliable as they were extensively cited by other reliable sources. The traditional media don't have a monopoly on reliability. Alaexis_¿question? 14:50, 14 May 2021 (UTC)

  This assessment is inapt and grossly unfair to Wikileaks. Broadly speaking, if Project Veritas publishes something then you can be confident that it is false and intentionally misleading; the same is not true for WL. (I say this as a person who agrees about the conclusion.) --JBL (talk) 15:49, 14 May 2021 (UTC)

• Option 2 WP:PRIMARY. There is some editorial control on WikiLeaks' end and more by various newspapers that collaborate with it. Obviously, not all documents have been verified by WikiLeaks or journalists, so WP:USEBYOTHERS is not absolute. But the bottom line is that editorial control exists to the degree it is possible in this type of publishing. As argued above, WikiLeaks has never been shown to publish false documents, so this editorial control has been effective. To wit, reliable but WP:PRIMARY.– Finnusertop (talk ⋅ contribs) 00:23, 14 May 2021 (UTC)
• Bad RFC - The RFC as presently constituted is not directed to a specific content issue. It is therefore impossible to give a decent answer. Wikileaks consists of a collection of primary sources of varying charactersitics and so the idea of rendering any kind of general view on its reliability is simply for the birds. FOARP (talk) 11:16, 14 May 2021 (UTC)
• Option 4 Wiki leaks does no independent verifying or factchecking, so they really should not be used at all.Jackattack1597 (talk) 18:32, 15 May 2021 (UTC)

  Do you have any evidence for that? Other people in the discussion have cited evidence to the contrary. Thryduulf (talk) 22:49, 15 May 2021 (UTC)

• Option 2, WikiLeaks is a collection of primary sources, and it should be treated as such. Devonian Wombat (talk) 10:22, 18 May 2021 (UTC)
• Option 2 There should be an evaluation of each Wikileaks document and be treated as a primary source. Sea Ane (talk) 12:28, 20 May 2021 (UTC)
• Option 1 Wikileaks verifies the authenticity of documents before releasing them, which accounts for sometimes lengthy delays between receiving and publishing them. There are no verified cases of any of the documents released being fraudulent. That is a higher standard than most reliable sources. That of course does not mean that that the information in the documents is necessarily accurate, since that depends on the original authors. TFD (talk) 12:58, 23 May 2021 (UTC)
• Option 3 or 4 per Aquillion, JBchrch, and MjolnirPants. Wikileaks itself is a collection of raw primary source documents (raising serious WP:OR issues when used directly), and the organization itself has questionable reliability and processes. Any information from them should be cited though reliable journalistic sources (i.e. not-Wikileaks), which could be counted on to do their own fact checking. - GretLomborg (talk) 06:03, 24 May 2021 (UTC)
• Option 1-2 Wikileaks is a reliable repository of authentic leaked government and business documents. Wikileaks has an extensive partnership with the best journalists in the world. The United States diplomatic cables leak for instance are genuine US government embassy reports. The usage of the different stored documents needs to be decided on a case-by-case basis based on policies concerning the usage of government and business sources. --Guest2625 (talk) 10:25, 25 May 2021 (UTC)
• Option 2. Wikileaks does verify the authenticity of documents it publishes and many documents it has published has since been acknoweldged as authentic.. However I personally feel that Wikileaks in itself should only be used in the absence of another, more mainstream and widely accepted source on the same matter and statements by Wikileaks should be attributed otherwise.RedAlert 007 (talk) 12:01, 25 May 2021 (UTC)
• Option 1 and 2 - Generally, they are reliable. I think that if the information can be verified using a different more reliable source it should. But unless there is reason to doubt the document I would take it as accurate. Not the policy for here but I tend to apply WP:AGF to new outlets as well. Also, I would ask Wikipedia to consider a policy on using documents gathered using FOIA requests. DoctorTexan (talk) 06:59, 17 May 2021 (UTC) (Moved to this section from discussion by Alaexis_¿question? 09:15, 27 May 2021 (UTC))
• Option 2 I can't think of a single case where a document released by wikileaks has been fabricated which rules out 3 and 4 for me, however due to the sensitive nature of the material they often release editors should be extra judicious in their use of the source and use RS analyzing the releases if possible. Additionally, attribution should always be required. BSMRD (talk) 03:57, 29 May 2021 (UTC)
• Option 2 It might be reliable on some cases, but it acts as an primary source, and therefore it should be cited with related reliable news sources if necessary. Ahmetlii (talk) 20:37, 30 May 2021 (UTC)
• Option 4 Julian Assange's controversial remarks about the Murder of Seth Rich should disqualify Wikileaks from being used as a source for anything. It's unclear whether WL has any independent editorial control outside of Assange. Geogene (talk) 01:14, 3 June 2021 (UTC)
• Option 1-2 I'm unware of Wikileaks of ever fabricating government documents and does its best to verify the authenticity of documents before it releases them, hence 1. But I can understand how it would be treated as a WP:PRIMARY where then I'd lean towards 2. Canadianr0ckstar2000 (talk) 19:12, 3 June 2021 (UTC)
• Option 2 or basically fine as long as used with attribution. I'm especially not seeing much justification for Option 4, which is presumably near-deprecation, as Nils Melzer of the UN Special Rapporteur on Torture etc. (often cited as an expert/significant opinion) considered Assange to have 'never hacked, stolen or published false information, nor caused reputational harm through any personal misconduct'. The debates on reliability look politicised, as in most of the criticism comes from the countries that have something to lose from the leaks. And even then, there are still US outlets that would vouch for its authenticity. Donkey Hot-day (talk) 10:25, 6 June 2021 (UTC)
• Option 1; qualified that Wikileaks is a publisher of information and "reliability" means that Wikileaks doesn't lie about or alter the contents of what they distribute and that they generally publish stolen primary sources. The fact a document has been stolen doesn't mean it's unreliable, but it means we should use extra care when citing it. Likewise when the documents they publish are primary sources. When Wikileaks doxxes people we shouldn't use it as a source. We shouldn't be giving it WP:UNDUE weight above other primary sources either. And the fact Wikileaks is reliable does not mean the documents they publish are actually reliable. If Wikileaks publishes a Chinese govt source talking about how the Uyghurs are all super dangerous terrorists, that's not going to be a reliable source for the claim that the Uyghurs are all super dangerous terrorists as the Chinese govt is unreliable there. I'd also qualify that stuff from the really early days of Wikileaks when they were actually a Wiki isn't reliable.— Preceding unsigned comment added by Chess (talk • contribs) 04:21, 9 June 2021 (UTC)
• Options 1, there have been numerous claims made here that Wikileaks do not have any "fact-checking", yet no backing up of this notion has been provided. Other people are raising issues about WP:UNDUE, which I do not believe are relevant because we are talking about the reliability of the source, not due weight. As for my take on the issue, I believe that Wikileaks is generally reliable. Wikileaks, their publisher and their journalists have won multiple journalistic awards and claim a "perfect in document authentication", which no one here has been able to disprove so far, despite many claims. CPCEnjoyer (talk) 12:23, 12 June 2021 (UTC)
• Option 3 If something in Wikileaks is good enough it would be reported in RS. Treating a repository of allegedly authentic stolen documents will only feed the rightgreatwrongers and conspiracy theorists. -Indy beetle (talk) 22:34, 13 June 2021 (UTC)
• Option 1 or 2, the Italian first class RS treat Wikileaks with great respect. I am not aware of any glaring cases of data falsification or anything of this kind.--Mhorg (talk) 13:04, 15 June 2021 (UTC)
• Option 2 or 3 maybe even Option 4, this isn't a question of whether Wikileaks is itself a reliable source. It's not a source to begin with but a repository of primary sources. Regardless of whether the material it publicises are authentic or not, on its own it's not usable. News organisations with a reputation for fact checking and editorial oversight might use them as a source but that does not mean we should too. They could be used as an additional citation alongside a news report which discusses its material although that might be pretty redundant. If they publicise something that isn't covered by any reputable news organisation then that material lacks both due weight for inclusion or an adequate filtering process. Tayi Arajakate _Talk 15:22, 15 June 2021 (UTC)
• Option 2. Wikileaks is reliable as a repository or republisher of primary sources. Whether those primary sources are themselves accurate is a different question, and needs to be assessed by individual case. It is very well established that Wikileaks is highly reliable as a source for the content of the primary source documents it republishes. In many cases, it would be used with attribution ("According to an internal US state department cable..."). Cambial foliage❧ 15:26, 15 June 2021 (UTC)
• Option 4 for the purposes of selecting one of these options, but I'm in full agreement with comments by Rhododendrites, Aquillion, JBchrch et al. that this is a misdirected RfC. We're asking whether Dropbox-with-an-agenda is reliable, which on face should be answered no. We should be waiting for these primary sources to be used by other independent news orgs with better resources for fact-checking, etc. The well-known concerns about Wikileaks being compromised only add to this. Alyo (chat·edits) 16:04, 15 June 2021 (UTC)
• None Wikileaks should be considered a primary source, so our usual 4 categories do not apply. I personally like citing some primary sources as complementary information for the reader. That is, right after a reliable secondary source discussing the document is cited. Otherwise, citing random Wikileaks document is generally not appropriate. MarioGom (talk) 16:52, 18 June 2021 (UTC)

Voting (question 2) (Wikileaks)

• Yes, but not outside Option 2 I would be even more careful when citing opinions on subjects few people have idea about (that may significantly influence perception of the article and we will probably not hear the other side if the issue is contentious but local in nature). The source should be used, but particular caution must be exercised while citing it, except for non-controversial facts which can be cited as they are presented in the document. Better this kind of source than no source at all. Szmenderowiecki (talk) 04:53, 12 May 2021 (UTC)
• It depends. Where coverage exists in a clearly reliable secondary source, this is obviously preferred, but where it doesn't (and mention is still WP:DUE) then the circumstances need to be evaluated individually - why is there no secondary source coverage? Is the material plausible? Is there any evidence the material is incorrect? These questions need to be evaluated based on the original source, the reliability or otherwise of Wikileaks will in most cases be irrelevant. Thryduulf (talk) 10:23, 12 May 2021 (UTC)
• Yes, in most cases There is a strong preference with Wikileaks for additional sources to establish weight and firmly anchor the article content in the published literature, and a must for anything controversial, but in some areas where English language sources are lacking (such as the domestic politics of Laos) it's not a violation of sourcing guidelines to use Wikileaks to fill in non-controversial facts. Spudlace (talk) 10:48, 12 May 2021 (UTC)
• Obviously not. The question is self-defeating: by definition, if the information cannot be found in a reliable source, it should not be included on Wikipedia at all. Using unreliable sources is not an effective strategy to globalise wikipedia. In fact, it's even worse than having biases. JBchrch (talk) 12:19, 12 May 2021 (UTC)
• Not sure I understand where Question 2 is going, but basically I said what I think is the same thing in my original vote above: Wikileaks should ONLY be when vetted by actual reliable sources, should never be the first or only source for anything. --Jayron32 12:38, 12 May 2021 (UTC)

  The intent and context of the question was that Lao domestic politics (and their Kremlinology in particular) is a topic very scarcely, if at all, covered by RS, and Western RS in particular, so corroboration by RS would be not possible because the outlets simply don't cover the region, even when for Laos, the event (Party Congress) is important. Access to Lao media is also limited, as the Internet in the country is very poorly developed and this is a Communist country with few civil liberties. The question goes: should we make an exception in this case and cite WikiLeaks under some special conditions that differ from the answer in question 1 (which is a general answer) because of an objective lack of RS coverage which is caused by an event happening in an isolated country with little interest in its news? (This is the reason I have inserted the WP:BIAS link)

  This question should not be interpreted as whether to grant a waiver to cite any claim or fact asserted in a WikiLeaks cable and for which coverage in RS would be likely ample were the fact significant enough. This is what you answer in question 1, where you choose your default option to treat WikiLeaks; question 2 concerns a very specific situation. Szmenderowiecki (talk) 15:50, 12 May 2021 (UTC)

Wikipedia:Systemic bias does not lay out a “reasonable editor” standard in this regard, what do you mean by that? Horse Eye's Back (talk) 19:46, 12 May 2021 (UTC)

Treating this as a kind of a court deliberation where each editor (and voter) is a judge of sorts, it's something that would be called reasonable person in common law court proceedings. In this particular case, anyone with some knowledge of Laos as well as the way Communist parties, diplomacy and media function qualifies as a reasonable editor (which I assume everybody writing here is); in general, a person with a reasonable knowledge of subject matter discussed is one. I didn't want to write "consensus" because I can't write "consensus" if we are about to establish it here. Also, WP:BIAS is only meant to indicate here that the fact is significant but coverage by RS is scarce, because it's Laos after all, not USA, Western Europe, Russia, Middle East, China or even North Korea, which is rather frequently mentioned in the media. Szmenderowiecki (talk) 21:02, 12 May 2021 (UTC)

The non-existence of any reliable sources on a topic is not a reason to fall back on unreliable sources. It just means that Wikipedia doesn't cover it. WP:BIAS is not fixed by lowering our standards. --Jayron32 13:49, 13 May 2021 (UTC)

• Absolutely not. Claims that are not covered in RSes are not suitable for inclusion in this project. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 16:17, 12 May 2021 (UTC)
• I agree with JBchrch, Jayron32, and ᛗᛁᛟᛚᚾᛁᚱPants. Incidentally, "does your answer change" is a terrible question since the same answer has different substantive meanings depending on the person answering it. --JBL (talk) 18:00, 12 May 2021 (UTC)
• Echoing Thryduulf. As for the "Is there any evidence the material is incorrect?" question — if RS say Wikileaks is boosting untruth in some way then would certainly weigh their view heavier than WL. --Chillabit (talk) 20:37, 12 May 2021 (UTC)
• Irrelevant. Stop it. Just stop it. See also WP:TRUTH and WP:UNDUE. Guy (help! - typo?) 21:21, 12 May 2021 (UTC)
• If "a reasonable editor may conclude that the coverage from RS is likely to be minimal or absent on the subject" then we should be even less inclined to use them as a source. If this question is implying that RS ignoring it should grant some kind of exception to allow us to use it, I wholeheartedly disagree. If there are no good sources covering something, we should not accept bad sources as a substitute to allow us to cover it, we should not cover it. NonReproBlue (talk) 01:42, 13 May 2021 (UTC)
• Irrelevant. My vote for Option 2 clearly states that I believe this should be taken into account, but they can still be cited in circumstances where an official government document/view on something is acceptable with a primary source. -bɜ:ʳkənhɪmez (User/say hi!) 03:53, 13 May 2021 (UTC)
• Sigh No as per MP. Only in death does duty end (talk) 07:27, 13 May 2021 (UTC)
• No, if more information cannot be found, the information from the Wikileaks document still needs to be attributed to the document in question, and Wikileaks would need to be mentioned as well, as the publisher. Contrary to some editors above, I do think that Wikileaks could be cited, even in a case where a journalist hasn't covered the document in question. In that case however, in-text attribution of the information both to Wikileaks and the document in question would be absolutely essential, since we're dealing with a primary source. -Darouet (talk) 15:17, 13 May 2021 (UTC)
• Trivially, tautologically no. NPOV means representing fairly, proportionately, and, as far as possible, without editorial bias, all the significant views that have been published by reliable sources on a topic. No RS, no coverage. That might leave us "biased" in various ways, but as Jayron32 said, lowering our standards doesn't fix anything. XOR'easter (talk) 18:43, 13 May 2021 (UTC)
• No. It seems kind of weird that we are even talking about this. We go to considerable lengths to rely only upon reliable sources and to limit use of primary and tertiary sources. Then it's suggested that, specifically because we don't have any information as to reliability, we're going to turn to an unreliable repository of primary sources as citable information? That seems twisted. John M Baker (talk) 20:14, 13 May 2021 (UTC)
• Irrelevant. WP:DUE issues are a separate matter from the initial question. – Finnusertop (talk ⋅ contribs) 00:24, 14 May 2021 (UTC)
• No It will be a form of bias. Sea Ane (talk) 12:28, 20 May 2021 (UTC)
• No Lack of coverage in secondary sources merely means that the information in the documents was not noteworthy. TFD (talk) 13:03, 23 May 2021 (UTC)
• Absolutely not. Wikileaks documents about events not covered by RS are the ones most likely to be unreliable, and the use of Wikileaks in this case would be unambiguously unacceptable WP:OR. - GretLomborg (talk) 06:08, 24 May 2021 (UTC)
• What?, an RS, any RS is better than the conclusions of a "reasonable editor". They don't all have the same biases. This just verges on asking whether we should allow original research on Wikipedia. Tayi Arajakate _Talk 15:30, 15 June 2021 (UTC)
• This question is self defeating and irrelevant, I voted for option 2 in the first question, however if consensus finds that Wikileaks is not an RS then Wikipeadia policy is clear on this, no RS, no coverage on Wikipedia, no exceptions. If it fails WP:V then should not be on Wikipedia at all, period. RedAlert 007 (talk) 03:40, 18 June 2021 (UTC)

Discussion (WikiLeaks RfC)

This is far too complex an RFC to be useful, FWIW. --Masem (t) 04:56, 12 May 2021 (UTC)

Well, actually I thought of making it as useful as possible by having the first four questions considered in order to answer the fifth. Question 5 is the most important of those - comments for the first four are auxiliary and I did not intend them to carry as much weight as for the fifth (which is the reason voting for question 5 appears first). Moreover, all of these questions surfaced at least once in that discussion alone, not to mention previous dozen or so in the archives. Some seemed to assert that WikiLeaks have 100% legit documents; there have been questions about verifiability, potential weaknesses and usage in particular contexts. Alone these questions would be pretty useless and an RfC on these would be odd. Besides, my understanding of the RfC process is that every participant is sort of a juror, and IRL they are asked several questions at once for them to evaluate evidence and arguments on each of them (1-4) to deliver a verdict (question 5); what I only wanted is to separate each discussion so that it could be easier to parse through it and sum it up when an uninvolved user closes the RfC.

I don't deny this is a difficult topic, but we would have to discuss it sooner or later. Szmenderowiecki (talk) 05:57, 12 May 2021 (UTC)

I don't understand the purpose of the auxiliary questions; they make me a bit uneasy in that they read like they're trying to dictate acceptable lines of argument or reasonings for the primary question. I would suggest removing them (and also sharply trimming the RFC just to ask what is currently question 5, with no further details beyond a link to the discussion that prompted this) - the primary question is what matters; allowing users to come up with and state their own reasoning for that is the entire purpose of an RFC. I don't think you intended to write a non-neutral RFC, but in general it's safest (and best) to stick to one easy, straightforward question. I would also omit the word "genuine" (it is begging the question), and just say something like "Is WikiLeaks reliable for publication of government documents? This was prompted by this discussion." --Aquillion (talk) 06:24, 12 May 2021 (UTC)

Question 2 is important. The RSP entry for Wikileaks mentions tampering and it gets hoisted as an argument every time there is a discussion on Wikileaks. Alaexis_¿question? 06:35, 12 May 2021 (UTC)

After some evaluation of the arguments, I think that yes, I'll change it as Aquillion (in most points) proposed, I admit it was too clumsy. I also post the last diff of the expanded RfC for reference, as I believe considering all of these questions is important so that they could sort of guide your decision; but of course I did not mean to suggest to vote one particular way - you are free to express and argue your opinions whichever way you wish to.

I will retain question 4, though, because that seems to be the question coming from that particular dispute. I believe answers to all the other questions may be incorporated into your justification, either in vote or discussion.

@Alaexis: you may want to change the content of your vote and your vote, now that the auxiliary questions have gone, and only two are here in place. I, for instance, incorporated some arguments from these into my vote. You didn't vote for the second question (which was question 4), so I did not include your answer. @Thryduulf: I have copied your comment under questions 3-4 under question 2 after reformulation - the comment itself has not been altered.

Sorry for the false start and all the mess it caused - I'll do better next time. Szmenderowiecki (talk) 10:27, 12 May 2021 (UTC)

There is enough precedent to believe it's trustworthy, but there is no hard proof for it and is mostly circumstantial, so it doesn't fit the rules strictly. This is generally how leaks go unless the originator of the leaks (I don't mean the leaker) admits to its veracity and of course that is never going to happen. You need to take into account too that diplomatic cables are essentially correspondence and might have mistakes themselves, so if information contained there is later proven to be false or inaccurate, that doesn't need to be because of any tampering on WikiLeaks' part, as the creators of these can be responsible for such innacuracies on their own. WP:NOR applies, of course. FelipeFritschF (talk) 07:27, 12 May 2021 (UTC)

For Question 1, Options 1 and 2 are also not mutually exclusive. WikiLeaks has a very strong record of verifying that documents are genuine, but additional considerations do apply (the documents themselves, while genuine, may express opinions, may be WP:PRIMARY, etc.). -Thucydides411 (talk) 17:31, 12 May 2021 (UTC)

Option 1 is meant to give full endorsement to the resource, on or close to the level of NYT, WSJ, WaPo, Associated Press, AFP etc., that you would cite without much reservations and doubts. Option 2 may be not mutually exclusive if you believe that the resource is generally reliable (option 1) but you'd still not use it because of some issues concerning bias (for example, just as we don't give full endorsement for political coverage on HuffPost but we consider it generally reliable otherwise); option 2 also encompasses cases when you believe that we should only cite WikiLeaks for some types of coverage and not others (e.g. reliable for uncontroversial statements of fact, unreliable for the rest). That we should handle opinions and primary sources according to current Wikipedia policies is self-evident, so I don't believe it should be a factor in voting. That is at least the meaning I intended to put into the options. Szmenderowiecki (talk) 17:49, 12 May 2021 (UTC)

• What does "genuine government documents” mean in question 1? I’m assuming that means published by wikipedia but authenticated by an independent reliable source which is not wikileaks like BBC, NYT, etc? Horse Eye's Back (talk) 19:43, 12 May 2021 (UTC)

I think that for Question 1, Option 1 is supposed to mean that documents published by WikiLeaks can be assumed to be genuine. However, this is not how I (or it seems anyone else) has interpreted the question. The problem is that even if WikiLeaks does a good job of validating documents (as I believe they do, based on their apparently spotless track record), additional considerations apply, because the documents themselves may be WP:PRIMARY, may contain opinions, etc. -Thucydides411 (talk) 20:37, 12 May 2021 (UTC)

Sorry if I wasn't precise enough, but that is what I meant here. Option 1 would mean that if you say this is a diplomatic cable and it was published on Wikileaks, you know it's genuine by virtue of being published on WikiLeaks, and thus WikiLeaks is (generally) reliable for publishing these documents word-for-word. Additional considerations apply should not refer to standard Wikipedia policy arguments, because everyone should follow the guidelines by default - this RfC is not about whether to follow guidelines or to change them (at most we can discuss which in this particular case have priority). Generally it is meant to restrict the usage of the resource to specific areas (which you mention in your vote, e.g. not in BLP or in uncontroversial settings only). Szmenderowiecki (talk) 21:20, 12 May 2021 (UTC)

I guess I don’t understand why the question has “genuine” in it then, no matter which option you pick you are assuming that the document is genuine based on the question asked. Option 4 would still be under the presumption that the document is in fact genuine. Horse Eye's Back (talk) 21:39, 12 May 2021 (UTC)

The word "genuine" was included because there was (and is) a substantial share of editors who argued these documents are not genuine and/or impossible to verify whether they are genuine and therefore reasonable doubts could be raised on their authenticity, which is one of the main concerns raised in discussions on the topic. Contrary to your suggestion, the word "genuine" does not presuppose my attitude to these documents. Yes, you can believe the documents to be authentic but vote to declare the resource generally unreliable or deprecate it nevertheless, which seems what ᛗᛁᛟᛚᚾᛁᚱPants did; you can just as well believe the documents are not genuine and vote for option 3 or 4 based on that (and, if I were to vote for Option 4, odds are that I would mean exactly that). It's up to you to decide whether these arguments are convincing enough for you and argue them in the voting section and here.

You are right, however, when saying that I assume the documents to be genuine until proven forgeries or at least when there is reasonable doubt as to whether they are indeed authentic. This is a matter of principle for me - just as I assume all editors do their job in good faith, so I do with journalists, writers, and scientists, just until I stumble upon glaring errors, logical fallacies or blatant lies. It is also my belief that so far the concerns about integrity of WikiLeaks mentioned in the relevant Wikipedia article as well as here are yet to materialise, so I don't think there should be reasonable doubts, at least for now. But again, if WikiLeaks is going to be caught for forging documents or being a conduit for forgeries on a massive scale, I will revise my opinion.

Hope this helps. Szmenderowiecki (talk) 22:44, 12 May 2021 (UTC)

The “you” there is general not a reference to you Szmenderowiecki. Specifically I’m the one trying to figure out how to vote on this. The question itself presumes the documents are genuine, we are asked to consider a theoretical situation in which wikileaks publishes a genuine government document not a theoretical situation in which wikileaks publishes a document which may or may not be genuine. It seems like it builds on a prerequisite, which if I look at the original format of the question appears to be because it did. It seems that as is we have at best a leading question. Horse Eye's Back (talk) 22:51, 12 May 2021 (UTC)

If you struggle with your answer, you can remove the word "genuine". I did not write the question(s) to presuppose authenticity of the documents, because this is contentious in the first place. While it was indeed one of the leading questions in the previous version of RfC, after reformulation, I tried to strip it from its previous role (given two users have at once suggested the RfC needs rewriting) and tried to construe it as broadly as possible. In other words, do not automatically assume authenticity, just imagine you are presented with a reference which directs to a WikiLeaks cable, that's it. Szmenderowiecki (talk) 23:19, 12 May 2021 (UTC)

The problem is that removing the word genuine completely changes my answer, those are not comparable questions. In other words, what you’re saying here and the question that was originally asked don’t line up, they’re not the same question. “Genuine” does in fact require us to "automatically assume authenticity.” The current question does in fact presuppose the authenticity of the documents. Theres no way around that without re-writing the question, I’m sorry if you didn’t ask the question you meant to but we can’t really change that now. Horse Eye's Back (talk) 16:08, 13 May 2021 (UTC)

Masem, actually, I think it is rather simple. You may well disagree, but for me, the fact that this debate exists because editors were unable, or refused, to find any unambiguously reliable source that include the information, says it all. Guy (help! - typo?) 21:24, 12 May 2021 (UTC)

Masem was commenting on an earlier version of the RFC that asked multiple questions (ie. it was too complex structurally, not too complex in terms of the core underlying issue.) --Aquillion (talk) 21:31, 12 May 2021 (UTC)

Yes, this was the state of the RFC my comment was directed to. --Masem (t) 22:04, 12 May 2021 (UTC)

Masem, Ah, ok. Well, WP:FUCKTHATNOISE covers the core issue, for me ,so. Guy (help! - typo?) 22:20, 12 May 2021 (UTC)

• @NonReproBlue, @Guy: Just to make sure - we are not evaluating reliability of Julian Assange and whether his views are WP:DUE and admissible (which may belong to the article about him but certainly not to this discussion), so any comments about what he thought of the documents and the conclusions he has drawn from the documents are not relevant. The fact that third-party bad-faith actors (and Assange himself) used the documents in an ugliest way possible, i.e. to create conspiracy theories, fake news and make unsubstantiated allegations doesn't mean that the documents themselves have been manipulated or doctored; even the fact Assange publicly lied about the source of the document does not mean the documents were not verified beforehand or not published unaltered. Actually, your statements that Assange was driven by his agenda and conclusions from the Mueller investigation (Assange must have known that Rich could not have been the source of the leak, because he received the mails when Rich was already dead and continued to confer with the Russian hackers to coordinate the release of the material., quoted from Julian Assange), prove the opposite - he knew the true source of the documents, he cooperated with Russian hackers, so there must have been at least some review before the documents were published, in this case by Assange himself. Now that the documents have been verified, the question stays whether they were altered, and by all indications they weren't, because nobody credibly suggested the documents themselves were fake.

I agree that multiple security experts warned against using the documents at face value, but I again heard no such expert saying that this particular document was fabricated or altered (and they should be the ones who are closer to the tools to verify the information), so in my opinion, this is so far a theoretical possibility, which should be taken into account when citing the resource (if allowed to use) but should not serve as an excuse to blanket ban the documents, whatever their content.Szmenderowiecki (talk) 03:34, 13 May 2021 (UTC)

Szmenderowiecki, no, we are discussing whether stolen primary documents can be crowbarred into Wikipedia despite the general unreliability of Wikileaks. And the answer is: no. Sources need to be reliable, independent, and secondary. Guy (help! - typo?) 07:12, 13 May 2021 (UTC)

On what policy is this assertion based? It seems to contradict WP:PRIMARY which says that primary sources can be used in certain cases. Regarding the independence, biased sources are expressly allowed. Alaexis_¿question? 07:31, 13 May 2021 (UTC)

That's your way to frame it. I don't want to "crowbar" documents despite consensus (which does not exist so far, otherwise there would be no RfC), I only politely ask if they are admissible, and if the consensus emerges the documents should not be cited, so be it. Contrary to your assertions that users supportive of using WikiLeaks are necessarily "Assangites" and insinuations they are acting in bad faith, they (we) are neither. I understand your opinion on WikiLeaks is that it is unreliable; the purpose of that comment, however, was to show that at least some parts of your argument are, in my opinion, flawed, and probably turn your attention to them. Szmenderowiecki (talk) 07:43, 13 May 2021 (UTC)

This might be a good time to remind people that the New York Times also publishes "stolen primary documents" (more commonly known as "leaked documents"). In fact, one of the most famous episodes in the paper's history was the publication of a "stolen primary document", the Pentagon Papers. As far as Wikipedia WP:RS policy goes, whether or not a document was leaked is irrelevant. What we're discussing here is whether WikiLeaks validates the documents it publishes, and it appears that WikiLeaks has a very strong track record of doing so. Its major publications are widely considered genuine, and nobody here has yet provided any examples of WikiLeaks publishing fake documents. -Thucydides411 (talk) 20:56, 13 May 2021 (UTC)

Intentionally lying about your sources in order to push a conspiracy theory is the kind of behaviour that is absolutely disqualifying as an RS. They are not a reliable source. At very best they are a collection of possibly genuine, selectively released primary source documents. If reliable sources cover something they leak, we can cover what they say about it. Otherwise we shouldn't cover it at all, just like any other document of unknown provenance or authenticity. Without RS covering a document contained in a leak, it absolutely fails the standard of due weight. We cover things in proportion to their coverage in reliable sources. If reliable sources give it zero coverage, then that is the same proportion we give it. No information is "important" enough to justify including it when reliable sources don't cover it.NonReproBlue (talk) 08:22, 13 May 2021 (UTC)

I believe that undermines the credibility of Assange but not the authenticity of documents. No one says we can't correct the source if RS unanimously say the source is different. Szmenderowiecki (talk) 00:26, 14 May 2021 (UTC)

• BTW, I believe editors participating in the discussion may find this table useful:

Sources: 1 2 3 4 5 6 7

I conclude that a majority of courts makes at least some use of WikiLeaks in their rulings, but few explicitly allow such evidence to be entered. Szmenderowiecki (talk) 07:43, 13 May 2021 (UTC)

It's also worth noting that the courts would be considering the reliability and/or admissibility of the specific documents relevant to the case at hand, not the reliability and/or admissibility of documents from Wikileaks as a whole. It is possible for different documents made available by Wikileaks to be differently reliable. Thryduulf (talk) 11:12, 13 May 2021 (UTC)

Well, some did. Special Tribunal of Lebanon and Court of Justice of the European Union have addressed the issue directly (whether it is admissible in general). And while indeed most of these rulings concerned particular applications and particular documents, the fact that a majority of the courts drew from the WikiLeaks cables while providing their reasoning to the judgment suggests that majority believes them to be authentic, and WikiLeaks reliable. Citing shoddy documents undermines the credibility of the court and is a very good case for appeal/rehearing, which the judges understand, so they must have evaluated their reliability, authenticity as well as conformance with current laws and bylaws concerning the procedure of admission of previously illegally obtained evidence before citing it or at least relying on it to issue the verdict.

Of course, quality of material dumped on WikiLeaks may be variable, so it might be that other courts, given the same documents, could reach other conclusions, but that's the current picture. Szmenderowiecki (talk) 13:21, 13 May 2021 (UTC)

And so what? Courts also use unpublished oral testimonies as their main sources to decide cases. So can I use an unpublished oral testimony as a source on Wikipedia now? Obviously not. The judicial process and wikipedia are completely different processes, with diametrically different aims and methods. JBchrch (talk) 12:40, 13 May 2021 (UTC)

My point was not to equate Wikipedia to a judicial procedure. That said, just as the judiciary, we have to evaluate whether the evidence is reliable enough, verifiable and authentic to be admissible. In this way, Wikipedia and the courts are fairly similar. And anyway, if I can file a lawsuit and win it based on cables obtained on WikiLeaks, it speaks volumes about the quality of the resource; conversely, if the courts consistently declined to view my claims based on WikiLeaks revelations or if plaintiffs/defendants who relied on these consistently lost their lawsuits for non-technical reasons, it would be a good indicator not to use it on Wikipedia. Here, the record is slightly in favour of WikiLeaks - I could not find more papers or news concerning WikiLeaks admissibility, so I think that's the full picture as we have it now. Szmenderowiecki (talk)

It's actually hard to provide a full answer to this comment because the reality is so much more complex then you try to portray it. So here are just two high level comments. First: As a matter of principle, courts accept everything into evidence: handwritten notes, UN reports, blood stained shirts, press releases by the US Department of State, used condoms, text messages, bags of trash... The fact that something was accepted into evidence indicates nothing about its reliability. Second: Most often, the question of admissibility is not related to the material reliability of the piece of evidence in question but to the question of whether it was illegally obtained. And often, you find yourself in the possession of a highly reliable piece of evidence, which was unfortunately illegally obtained (classic example: a hidden camera footage of a private meeting). So I reiterate my point: admissibility in court and WP:RS are completely unrelated. JBchrch (talk) 14:22, 13 May 2021 (UTC)

To the first point: I can't agree with you, because I have explicitly cited cases from Cambodia and Lebanon which did not allow the documents to be introduced into evidence because they had doubts over their integrity and reliability; on the other hand, CJEU and UK Supreme Court endorsed WikiLeaks, so no, it's not automatic and it's not everything.

To the second point: All of the courts mentioned dealt with documents that were previously obtained against the law, and none of them dismissed the documents because they were illegally obtained some time before plaintiffs/defendants used them. Citing cases where Wikileaks documents were dismissed because they were illegal in the first place would be useless, because in these cases, reliability, veracity, authenticity etc. are not considered at all. Szmenderowiecki (talk) 14:39, 13 May 2021 (UTC)

First point: this is why I said as a matter of principle, i.e. there are exceptions.

Second point: All admissibility decisions are useless, because the standards they apply—may they be illegality or patent unreliability (which is, for the record, a way lower standard than the one we apply here)—has nothing in common with the standards we are supposed to apply. JBchrch (talk) 14:58, 13 May 2021 (UTC)

To the first point: sorry, haven't noticed these words. To the second, actually, if not in all, in most of the cases, the ruling itself was not specifically over whether to admit WikiLeaks but they mentioned it in a few paragraphs. Moreover, it seems that the outcome of most of the litigations mentioned hinged on whether WikiLeaks documents were admitted or not. In the UK case, it actually meant Chagos Islanders won against UK (because basically that was the main evidence of malfeasance and intent of the UK and US officials), so they must have investigated the document thoroughly. The reliance was not that large in Yukos v. Russia, Caratube Int'l Oil Company v. Kazakhstan and CJEU cases, but was still pretty substantial. The same can be said of cases where the WikiLeaks documents were dismissed as unreliable/impossible to verify their authenticity. I believe all of these cases are relevant; and your conclusions may be different based on the table - my, sort of, duty as OP of the RfC was to provide available evidence for community evaluation to make a better decision. Szmenderowiecki (talk) 18:06, 13 May 2021 (UTC)

I still fail to see how any of this translates to Wikipedia-reliability. Courts have the tools, the time and the ressources to analyse documents of questionable origins and any other dubious stuff the parties usually throw at them. We don't have that. In fact, Wikipedia is specifically built around the idea that editors should not do that. If you need citations regarding these affairs, then you can cite the court case or, better yet, a secondary source about the ruling. JBchrch (talk) 18:20, 15 May 2021 (UTC)

I don't say we should be engaging in WP:OR, as the policy says we shouldn't. What I say is that, from the resources I've dug out from Google on the subject on WL reliability/admissibility as evidence in courts, the majority either explicitly says it is or that it uses the documents to draft their opinions (and they must be impartial while doing it). Had the courts been unanimous in their rulings concerning reliability or at least unanimously used the resource to draft their rulings, I'd vote for option 1, but since it's only a majority, I opt for Option 2, and I specified that we should avoid drawing statements from WikiLeaks to Wikipedia if the matter is a subject of controversy, but for documents that are not (and are rather unlikely to cause it), i.e. for the category of documents that don't need OR to be determined faithful and authentic, I see no obstacles doing so.

The court cases are cited for reference in the table, you may check the details for each court case if you want; I added some names so that people could search them. Also, you have seven secondary sources that interpret them (and other original cases); I believe it will be fine for your analysis should you need it.

As an aside, I should note that international courts (and, apart from UK Supreme Courts, all of these are international), apply much stricter standards of admissibility than your local court you will normally sue anyone in, common law or civil law. Which is one of the reasons international courts have pre-trials and trials lasting several years. EU courts are largely civil-law ones, and they too seem to have a higher bar for admission of evidence than EU member state courts (unlike in US, where a lot of states copy federal guidance on admitting evidence) Szmenderowiecki (talk) 19:58, 15 May 2021 (UTC)

I really want to WP:AGF here, but you make it harder and harder, honestly. Regarding the length of international procedures, I assume that you are confusing admissibility as a question of jurisdiction (i.e. is the court competent to rule on this matter?) and admissibility of evidence? I am not aware that admissibility of evidence takes this much energy at international courts. However, I know that admissibility as a matter of competence is always heavily challenged by the parties, and it is in fact the topic of the first big decision in international criminal law. Regarding admissibility of evidence by international criminal law, the relevant literature says the following, which completely contradicts what you said:

Regimes as to the admission of evidence differ. Common law systems often have strict technical rules on the admissibility of evidence. They are meant to exclude irrelevant evidence, safeguard the rights of the Defence and protect a jury from exposure to unreliable or unfairly prejudicial evidence. Inquisitorial systems have a more liberal regime. They place more weight on the ‘free evaluation of evidence’. All evidence is generally admitted, and then evaluated by judges. This flexible approach is reflected in international criminal procedures. Procedural instruments grant judges a wide degree of discretion to rule on the admissibility of evidence. The idea is that evidence should be weighed at trial, rather than precluded per se. This approach takes into account the difficult context of international criminal investi- gations, including limited access to documentary evidence and witnesses. It is increasingly important in light of the multiplication of fact-finding and evidence-gathering bodies, and the absence of a single set of procedural rules governing investigations and prosecutions. It makes the acceptance of material as evidence dependent on the judgment of those who receive it.

Stahn, Carsten (2019). A Critical Introduction to International Criminal Law. Cambridge University Press. p. 343.

The approach to evidence at the Tribunals has been described as flexible, liberal and unhindered by technical rules found in national and particularly common law systems. Professional judges try both fact and law and there is no need to protect jurors from lay prejudice. The same is true for the ICC.

Cryer, Robert; et al. (2010). An Introduction to International Criminal Law and Procedure. Cambridge University Press. p. 465.

JBchrch (talk) 12:24, 16 May 2021 (UTC)

OK, these were more of my impressions from reading separate decisions about whether to admit documents into evidence or not, of which I'm not very much aware in common-law procedures and probably just a little in civil law; if scholars say admissibility is indeed a rather liberal procedure, I'm not here to dispute it :); though the fact the documents are frequently contested and the courts have separate decisions on each batch of evidence compensates somewhat for the laxity. My bad, it wasn't intentional. I am sure though that I don't mention admissibility as a matter of jurisdiction, because, from my reading, no court said it would not admit the documents into evidence because it couldn't rule on it, all that did rule did so on the merits. I wouldn't want cases on lack of jurisdiction anyway to be mentioned here because they don't rule on the contents of the resources.

OK, let's even suppose we don't take admissibility too seriously. My point is that if the judges use the reasoning provided in cables in their rulings, and by your admission, the judges have the tools and time to verify if the evidence is reliable and authentic, that means they established that the source is good enough to be relied upon, even if they don't rule explicitly on admissiblity or reliability. The corollary also holds true that if a court explicitly dismisses WikiLeaks or has a long practice of not mentioning the (alleged) facts presented from the evidence in WikiLeaks (which can't be said from here because no court has a long enough history of deciding on WikiLeaks), it should make us suspicious to use it. I still find the balance favorable for WikiLeaks, even when excluding strictly admissibility questions: Supreme Court UK, SCSL, 3 rulings of international arbitration decision, CJEU (2 cases) vs. STL, ICTY, (probably) 1 international arbitration decision and ECCC. Szmenderowiecki (talk) 15:00, 16 May 2021 (UTC)

More than just authenticity of the documents comes into play when deciding whether evidence is admissible in court. There are additional considerations that have nothing to do with authenticity that may prevent documents published by WikiLeaks from being admitted as evidence. In the Chagos Islanders case in the UK, for example, the UK Supreme Court had to consider the argument that admitting the cables into evidence would breach the Vienna Convention of 1961, which establishes the inviolability of diplomatic correspondence. The UK Supreme Court eventually ruled the cables to be admissible (), but this at least shows that considerations beyond authenticity can prevent documents from being admitted as evidence. -Thucydides411 (talk) 19:48, 17 May 2021 (UTC)

What you say is true, but I'll repeat again, cases where WikiLeaks has been dismissed on procedural/technical grounds are not mentioned here, and in particular no court has ordered the evidence dismissed/admitted while applying the Vienna Convention; I did mention some cases where they just ruled them admissible but nothing beyond that, which JBchrch suggests we should also not take into consideration, and he might have a point if the evidence was admitted and no one made any specific remarks on the resource's quality.

Btw, the resource you cite is mentioned as number 2 above the table. Szmenderowiecki (talk) 16:26, 18 May 2021 (UTC)

• The problem is that we have no way of knowing whether the version of a document hosted on Wikileaks is a “true and accurate copy” of the original - or whether it has been tampered with.

Eventually, the government will release the original document to the public, and at THAT point we can compare it to what is hosted on Wikileaks. IF there are no discrepancies, THEN we can cite the original and link to the version hosted on Wikileaks (as a “convenience link”). Until then, no. Blueboar (talk) 11:31, 13 May 2021 (UTC)

That's a selective application of an impossible standard of accuracy, in spite of no evidence of actual tampering. Also, they do have a verification process, as noted by Burrobert earlier . Alaexis_¿question? 12:27, 13 May 2021 (UTC)

Not an impossible standard, just severely limited. The original document will (eventually) be released and thus citable... and (in most cases) Wikileaks can then be used to view it. Just not YET. Blueboar (talk) 12:52, 13 May 2021 (UTC)

• Szmenderowiecki, no, the list of cases is not particularly helpful. Wikipedia is not a court. The question for Wikipedia is whether we should cite stolen copies of primary documents hosted on a website with a clear political agenda and considered, in our terms, of questionable reliability at best.

As Wikipedia policy questions go, that's about as simple as you can get: No. Sources are supposed to meet the trifecta of reliable, independent and secondary, and we must not give undue weight to things.

If the fact is true and not contained in other sources, it is not significant.

If it is true and contained in other sources, we use them instead. Guy (help! - typo?) 13:48, 13 May 2021 (UTC)

1. I hope you read the explanation as to why I listed the court cases. I know Wikipedia is not a court, but it does share some principles with the court system, one of which is to exclude evidence that is likely to be unreliable, forged, or both. And the courts dealt with evidence that was clearly obtained against the US and other laws, and argued in favour of those using the documents and/or used the documents themselves in a majority of cases. Please stop arguing that the document is not admissible because it was stolen X years ago - it's now on public domain and only Wikipedia policies may bar us from using it, which we are to determine here.

2. No policy on Wikipedia says the source must be all of three (and yes, even if you author an otherwise brilliant essay, policy guidelines are more important than essays). It must be reliable, agreed, no exceptions (that's to be decided). WikiLeaks, unlike regular outlets like NYT, does not produce news themselves and is only a repository of documents, as JBChrch rightly noted, so independence principle does not apply here, and even if it did, bias is not something that disqualifies the resource, whatever your opinion on Trump is. Verifiability, on the other hand, does, which I believe can be inferred from a clean record when it comes to documents per se (not how others interpret them). It needn't be secondary, otherwise WP:PRIMARY would be redundant. WP:PRIMARY expressly says primary sources may be cited, but we should be cautious. On the other hand, there's almost unanimous consent that, faced with the choice to cite WL or secondary RS, we should cite the latter. We don't always have that luxury, however, which was the case in the disputed description of a Lao politician. It does not follow automatically that the fact is not significant. Most Europeans or Americans would say "whatever" if the Chinese built another dam on the Mekong, but for Laos that's important, and that should be our vantage point. That attitude is the reason we can't find the news, not because Laos itself is insignificant (even if it is small and poor). Szmenderowiecki (talk) 14:39, 13 May 2021 (UTC)

Would wp:copy come into this?Slatersteven (talk) 13:52, 13 May 2021 (UTC)

• Can we please have an extended confirmed requirement for opening one of these RfCs? There are lots of them, and it's not always worth settling on which shade of lousy a source is. — Rhododendrites ^talk \\ 14:27, 13 May 2021 (UTC)

• Bobfrombrockley: To comment 1): There are four prongs of WP:USEBYOTHERS: 1. How accepted and high-quality reliable sources use a given source, 2. whether they are used often. 3. whether the coverage is positive or negative, and 4/3a. whether RS release information "as is" or heavily comment on it and its veracity. Even if we assumed Donald Trump is a source (even though WP:USEBYOTHERS concerns other media outlets in general, not person's opinions, but so be it for the purposes of the argument), we just say he's largely unreliable, because while he is covered by RS extensively, the coverage about him personally is negative in the majority of RS (particularly since late 2020) and the majority also comments extensively on his claims to rectify them. In general, though, what Trump says has much more to do with WP:OPINION, or, as in the case of 2020 election, WP:FRINGE.

Zaathras: I don't believe the comparison is correct. Project Veritas is known to repeatedly manipulate their videos which they purport are how it looks like IRL so that the impression from the dialogue is different from what you'd hear in full dialogue - there is no known instance the same happened with WikiLeaks's documents (redaction of which does not preclude authenticity). Then, unlike WikiLeaks, Project Veritas settled a libel lawsuit against an ACORN employee, in which the defendant admitted having created deceitful coverage, and that's only because common law allows settlements that they weren't indicted; WikiLeaks AFAIK was not subject to any. You also say they are not known to be fact-checking or verifying the documents, but sources submitted here so far indicate to the contrary. Szmenderowiecki (talk) 00:26, 14 May 2021 (UTC)

Your fevered opinions supporting Wikileaks are of no interest to me, thanks. Zaathras (talk) 01:15, 14 May 2021 (UTC)

Option 1 and 2 - Generally, they are reliable. I think that if the information can be verified using a different more reliable source it should. But unless there is reason to doubt the document I would take it as accurate. Not the policy for here but I tend to apply WP:AGF to new outlets as well. Also, I would ask Wikipedia to consider a policy on using documents gathered using FOIA requests. DoctorTexan (talk) 06:59, 17 May 2021 (UTC)

@DoctorTexan: You may want to move your comment to the "Voting for question 1 (WikiLeaks)" section above. -Thucydides411 (talk) 10:01, 17 May 2021 (UTC)

I would ask Wikipedia to consider a policy on using documents gathered using FOIA requests. Here you go. --JBL (talk) 19:35, 17 May 2021 (UTC)

DoctorTexan, I've copied your vote to the survey section. If this wasn't your intention, please remove it from there or let me know and I'll revert myself. Alaexis_¿question? 09:15, 27 May 2021 (UTC)

There has not been any question of the reliability of Wikileaks' documents from reputable observers. While some of the targets of Wikileaks have questioned the accuracy of the documents, none of them have provided any evidence.

I would caution against using any primary source that has no coverage in secondary sources, since it raises problems with original research and weight. It requires original research to interpret primary sources and if information does not appear in secondary sources, it lacks weight.

There was a similar RfC after Wikileaks released documents relating to the War in Iraq. You should provide a link.

TFD (talk) 13:13, 23 May 2021 (UTC)

Yes definitely agree with that view. The question of the reliability of the Wikileaks documents needs to be separated from the issue of when it is appropriate to use them. As with any primary document, we should not be introducing a Wikileaks document into an article without some good reason, e.g. when it has been covered by secondary sources. The process of choosing a particular document to cite, even if the document is presented without any interpretation, would generally involve original research. Burrobert (talk) 13:23, 23 May 2021 (UTC)