Regarding My Ban from Wikipedia

Since I have been contributing to Breitbart, there have been several comments made regarding the fact I was banned from Wikipedia. For the sake of clarity I am offering this description of the situation as it played out. A summary is included at the end of the piece. You can also refer to Breitbart’s own coverage of my situation in the following articles:

• “Wikipedia Can Now Ban You For What You Do On Other Websites” by Allum Bokhari on January 4, 2016
• “EXCLUSIVE: Banned Wikipedia Editor Says He Hasn’t Seen Any Evidence Against Him for ‘Harassment’” by Allum Bokhari on January 14, 2016
• “Banned Wikipedia Editor Asks Jimmy Wales Why He Was Removed From Site” by Tom Ciccotta on June 28, 2016
• “Wikipedia Elites Change Reasons for Banning Dissident Editor” by Allum Bokhari on March 24, 2017

Other coverage there has also referenced my situation:

• “Slate Columnist David Auerbach Criticized Wikipedia. Now Wikipedians Want Him Fired” by Allum Bokhari on July 5, 2016
• “Wikipedia’s Seven Worst Moments” by Charlie Nash on July 5, 2016

For the e-mails described below related to my ban, see here for the redacted versions. As it concerns the identity of the admin I reported to the Arbitration Committee, who banned me in response, see this statement on the only details I will disclose with regards to that admin. I would note my statement applied at the time I made it, but in the event that should no longer be an accurate description, I will not make any correction to that effect. Note: this piece may be updated should the situation with my ban develop further.

Some time back I looked into a public allegation made by another editor that an admin was editing on a matter where that administrator had an undisclosed conflict of interest. Upon investigating the matter, I learned the administrator’s likely identity and was able to confirm this admin had edited on that matter and various others where the administrator had an undisclosed conflict of interest. Most of this involved edits about close friends or colleagues, but some involved self-promotional edits as well. Of most concern was one instance where admin tools were used in violation of the policy regarding admins involved in a dispute.

Although these actions represented violations of the site’s general content policies, breach of administrative policies, and broke the policies and guidelines regarding conflict of interest (COI) editing, I avoided any action at the time. Initially, it was in part because I had only noticed one major dispute where this administrator had engaged in misconduct while having an undisclosed COI and a few other articles.

Later on the admin took another administrative action on one of the matters where I already knew that administrator had an undisclosed COI. I investigated and found another major dispute where the administrator had engaged in misconduct while having an undisclosed COI. This led me to find strong evidence that the administrator was who I believed as the person I believed to be the administrator announced the existence of a Wikipedia article the administrator had created mere hours beforehand.

As I had only learned of this long after the admin action had been taken and there were some sympathetic reasons for the action, I opted to instead inform the admin privately of my concerns and warned the administrator that I would go to ArbCom, who are charged with examining administrative misconduct, if there were repeat offenses of editing matters where the admin had an undisclosed conflict of interest. I never received a response, but for over a year that seemed to be effective at preventing this administrator’s misconduct. However, a change in my fortunes would undermine my efforts.

Following the badly misreported GamerGate ArbCom case on Wikipedia, I was barred from edits about the topic and gender-related disputes. Additionally, I was given general restrictions limiting the number of times I could revert an editor and received a formal warning saying I could be banned by motion if future misconduct occurred in any topic area.

The lead up to my ArbCom ban involved three different blocks by administrators whose conduct I had criticized during that case. Although there were issues with all three actions, my intention is not to elaborate on those incidents. My third block was for three months and also involved removing the ability to use my talk page. During this extended block, I was still able to send e-mails to other users through Wikipedia and did so on various matters.

One other apparent result of my block was the administrator returned to undisclosed COI activity. At first this was undoing changes made to an article that restored the admin’s own promotional editing. There were legitimate concerns regarding some of the edits, so I sent another warning with advice on how to undo the parts that restored promotional content, while keeping the valid removals. No action was taken to this effect, but some of this content was removed by another party and the effect was the same.

Unfortunately, the administrator did not stop and instead significantly expanded an article on a friend and made various edits about another friend. Still I ignored this until I saw some edits made to another article about someone who was involved in a criminal matter where there was arguably a conflict of interest. The edits included violations of the site’s policies on articles about living people, particularly those of people accused of crimes.

I would acknowledge this person is not the most sympathetic character, but there was history of similar editing by this administrator regarding another person where there were conflict of interest issues. Given I had already warned this admin twice to no avail and the behavior only continued, I was not about to wait and see if there would be a repeat of the same conduct with a more sympathetic target. My response was to give the admin a choice to either disclose the conflict of interest on various articles or be reported to ArbCom in a week. The admin apparently opted for the latter.

After my e-mail, something to further confirm the admin’s identity occurred. I had mentioned the friend whose page this administrator significantly expanded in my e-mail. On a page outside Wikipedia under the administrator’s real name where the administrator mentioned that friendship, the claim of this person being a friend was removed and replaced with milder language about this person’s connection to the administrator. This change occurred soon after my e-mail was sent.

Having no uncertainty regarding my position, I submitted my report to ArbCom laying out all of the above details with extensive diffs and other links to evidence confirming the identity of the administrator and connections to subjects the administrator edited about. Nearly all the details unrelated to the disputes that initially turned me on to this administrator’s actions were based off details in the Wikipedia article about the person I identified as the administrator, the rest being simple searches of the admin’s real name alongside the subject of any articles edited. Then I informed the administrator of the report and suggested the admin could disclose the conflict of interest and resign as an admin to avoid any sanction from ArbCom, which I expected would occur.

One thing I noted in this and to some extent in previous e-mails is that disclosure of a conflict of interest could lead to the administrator being outed by other parties, which is also something that has been noted in the COI guideline in various forms. At no time did I suggest I would reveal the administrator’s identity to anyone but ArbCom and I explicitly pledged not to reveal anything I knew in this last e-mail.

Within hours of my e-mail to ArbCom I received a response from Arbitrator GorillaWarfare stating:

This is to acknowledge receipt of your message, the contents of which have been circulated to the committee for consideration.

If we believe further information from you would be helpful, we will be in touch.

They never got in touch. A couple weeks later on New Year’s Eve I received a notification that I had been mentioned on ArbCom’s noticeboard. Clicking it, I saw this message was posted:

In remedy 8.5 of the GamerGate case, The Devil’s Advocate was ‘strongly warned that should future misconduct occur in any topic area, he may be banned from the English Wikipedia by motion of the Arbitration Committee.’ Accordingly, for continuing harassment of other editors, The Devil’s Advocate is banned indefinitely from the English Wikipedia. He may request reconsideration of the ban six months after this motion passes, and every six months thereafter.

Even though I suspected this was a response to my e-mail, one thing about the message caught me by surprise. The message claimed “continuing harassment of other editors” and yet if the ban concerned my report, it would not make sense to mention harassment of multiple editors. I immediately shot off an e-mail to ArbCom (redacted for identifying information):

I demand a serious explanation immediately for the cause of this accusation that I am harassing editors. At no point was I informed of any discussion about banning me or any accusations of harassment. The only conclusion I am forced to reach is that this concerns my e-mail regarding [redacted]. Warning an admin against abuse of [redacted] position and then informing to the proper authorities on Wikipedia about this abuse when it continues unabated is not harassment in any way, shape, or form. You are committing acts of libel against me if you are basing your decision on my communications with [redacted] and the Committee. Please clarify exactly what informs your decision and why you felt I didn’t need to be informed beforehand.

I sent several messages to ArbCom regarding comments their members were making about the ban that I felt were misrepresenting the nature of my ban and any claimed harassment, particularly a comment by Gamaliel suggesting I had offered some statement in my defense. Subsequently, an administrator who is married to one of the arbitrators who voted on the case had made a comment referencing my case that provided an analogy of a blocked editor who believed they were “whistleblowing” and threatening to dox admins.

After someone asked me about whether her comment was describing my situation and given the slight resemblance (including the mention of multiple editors due to ArbCom’s own misleading announcement), I contacted her to say I felt it was libelous and asked that she clarify it was hypothetical. She made the change and, despite having been recused due presumably to our being parties in the GamerGate dispute, Gamaliel blocked my access to e-mail on the site. One arbitrator, Doug Weller, boasted that as many new members of ArbCom that year were elected on an anti-harassment mandate that this proved they were fulfilling the community’s wishes.

During this time I also contacted various people who I had e-mailed over the course of my block to discern if any of them might be claiming harassment or if someone I discussed with them might have heard and reported it as harassment. I had suspicions that one or both of the recused arbitrators might have also claimed I was engaged in harassment of them as well. Nothing I did would qualify as harassment, but I set out to exclude possibilities and every last alternative explanation was excluded. ArbCom did moderate the ban message in the face of community criticism, stating that my actions violated the harassment policy and noted it concerned a single incident.

When I mentioned in my interview with Breitbart that I was not told by ArbCom what my ban concerned this sparked some further controversy on the site and members of the ArbCom ultimately confirmed this was the case. I had at this time noticed part of the policy for the Arbitration Committee relevant to my case:

In exceptional circumstances, typically where significant privacy, harassment or legal issues are involved, the Committee may hold a hearing in private. The parties will be notified of the private hearing and be given a reasonable opportunity to respond to what is said about them before a decision is made.

This policy is a community-mandated policy, which means the editors on Wikipedia are the only ones with the authority to substantially modify the policy. It offers no leeway for ArbCom to reinterpret the policy. Soon this was raised in the discussion about my ban as well. Two weeks went by without a response from anyone on the Committee before they all offered statements in their “personal capacity” that they believed ignoring this policy was necessary to protect the “victim” of harassment. Arbitrator Drmies went so far as to repeatedly call me an abuser. Weller stated he had asked the Wikimedia Foundation about the matter and claimed they approved of ArbCom ignoring the arbitration policy.

Finding no favorable response, I submitted an appeal to Wikipedia co-founder Jimmy Wales, who has the power to overturn ArbCom actions and sought to alert him to it on the Meta Wikimedia site. During this time, the administrator again breached the conflict of interest policy and I promptly notified ArbCom. After another editor was indefinitely blocked under similar circumstances in March the discussion reignited on Wikipedia. One editor suggested an amendment to the arbitration policy per the requirement for community ratification to expressly allow the approach taken by ArbCom and it failed by a resounding 25 to 1 vote. Yet another instance of questionable ArbCom secrecy led to me finding this discussion about ArbCom in which Wales gave a clear denunciation of the exact process used against me stating:

I think the important statement has been made: no secret trials, and no convictions without giving the opportunity to present a defense. That’s just basic justice, and I will overturn any ArbCom decision to the contrary. (Although, I should point out, there is ZERO chance of the ArbCom doing this in the first place.)

Given his pledge to overturn any such case, I pressed harder for a response and after it was raised on WikiInAction another user posted my comments at the English Wikipedia talk page for Wales. At this point, after months of evasion, Wales finally confirmed that my e-mail to ArbCom was indeed what precipitated my ban from Wikipedia. In the process, he proceeded to provide many false and misleading characterizations of the e-mail claiming “the harassment was confessed to in extensive detail by the now-banned user, in an email to ArbCom.”

Wales described my report as “7 pages of wall-to-wall text detailing a campaign of harassment over an extended period of time (over a year), including detailed investigations into another users personal life, repeated threats to that user to out him, ‘opposition research’, etc.” He stated I “testified extensively in a 7 page email detailing his own misconduct.” At that point Wales suggested I had “pretended” not to know the reason for my ban shortly after arbitrator Weller had suggested I was not notified to keep me from knowing the reason. Even in the face of all this, Wales stated ArbCom’s failure to notify me was wrong, but argued it was “within their discretion” to do so despite the community-mandated policy forbidding it and his own words in the past repudiating such a proceeding. He cited my warning remedy from the GamerGate case as the reason.

I responded with refutations of every point and argument he raised. One thing I noted was that another editor from the GamerGate case, Tarc, had an identical warning remedy placed on him. Unlike in my case, when the editor doxed an administrator on Twitter and engaged in pointless sexual mockery of the admin, he was allowed a chance to take part in the proceedings that led to his ban. Wales acknowledged my overall response to his arguments was factual, but that he differed on the interpretation of the facts. At this point, I posted redacted versions of the e-mails for all to review.

One user pointing this out criticized me for sending warnings before reporting, but suggested my conduct was still well within the rules. When challenged on the COI report, Wales insisted members of ArbCom reviewed the evidence and determined my report was “baseless” yet stated he did not review the evidence himself. Two arbitrators weighed in to drive home this interpretation with Drmies stating, “There was nothing to the complaint, nothing whatsoever. It was investigated, in detail; TDA’s charges are without merit, and their reading of the history and of policy are simply wrong” Opabinia regalis claimed Wales was quoting her own e-mail and affirmed it as her opinion.

Drmies even took the extra step of stating I was “wrong” directly to me, which was the only point in this discussion that any of them spoke to me rather than about me. In his response he stated “as a matter of fact, it seems that the subject strongly disagreed with the admin’s edits that you think were so friendly to the subject.” This caused me to question if Drmies was suggesting the admin was not who I identified to them. Further inquiry only affirmed what I had found as I discovered logged-out edits with an IP matching the known location of this person I identified as the admin. Some of the edits also revealed the administrator likely got started as a paid editor.

What I came to find over the weeks and months ahead is how truly extensive this administrator’s COI editing has been over the years. From what I was able to discern, the administrator has likely engaged in some significant improper editing on a matter where there was an undisclosed conflict of interest every year since the administrator started editing Wikipedia. Not only that, but I found arbitrators involved in my ban had also raised concerns about this administrator’s editing in the past and may even have been aware at least of the admin’s identity if not the conflicts of interest. More evidence amassed that the administrator was who I said, including an admission by the person in question to being an administrator on the site. All of this was forwarded to ArbCom. The administrator also continued COI editing after my ban, something I reported to ArbCom as well.

Alas, it has all proved to be to no avail. My first appeal where I was finally able to lay out a detailed defense was rejected. The arbitrator sending me the message stated they disagreed with my conclusions about the administrator and that my ban would not be lifted unless I agreed to not “explicitly or implicitly” threaten to “out” anyone. I was also told my ban would not be lifted unless I “dropped the matter” of the admin’s COI editing. This was the first official response I received from ArbCom regarding the reasons for my ban after over half a year. Given the strength of the evidence I presented both before and after my ban and the fact I did the opposite of threatening to out the administrator, I was not appreciative of this response.

Even so, despite my attempt at clarification on their terms going unheeded, I agreed in my second appeal, after the other six month period expired, to only contact ArbCom about this administrator’s editing and to not discuss it publicly any further. I also restated my commitment to following the outing policy, while stipulating I never violated it. Both times I rejected any suggestion of further restrictions being placed on my account, something I understood would be just as harmful to my appeal as my continuing to state my innocence.

Yet ArbCom did not simply reject my appeal for being short of the groveling false confession that they wanted, but instead changed arguments entirely. Gone were the claims of harassment and replaced with new claims of “tendentious editing” and the evidence for this were edits before and e-mails after my ban. It was not lost on me that e-mails before my ban, the entire basis of my ban, were now excluded from consideration in future appeals that would also be pushed back to a year from this latest rejection. Following Breitbart’s coverage of this, I posted redacted versions of the appeals and the responses I received.

Just as has been the case in apparently every year since the administrator began editing, there were more incidents of undisclosed COI editing. This year even saw a significant escalation in such editing and I have good reason to believe ArbCom is aware of at least some of this editing and knows it has been to some extent promotional in violation of the site’s policies and guidelines. As of yet there is no sign of sanctions being imposed on this administrator. Whether that will change any time in the future is anybody’s guess, though ArbCom’s response so far leaves little room for optimism.

TL;DR

To provide a summary of the matter:

• I learned some time ago that a Wikipedia administrator was editing and at times using admin tools on matters where the administrator had an undisclosed conflict of interest.
• I privately warned the administrator that I would speak to the Arbitration Committee, who are responsible for reviewing admin conduct, if the conduct continued. Once that proved fruitless and the admin’s behavior escalated, I sent the report.
• Rather than take action against the administrator, ArbCom banned me for “harassment” without any notice leaving me no chance to defend myself or even know the cause or reasoning for the ban.
• Editors at Wikipedia challenged ArbCom’s actions and when learning all the circumstances pointed out their failure to notify me violated binding policies imposed on ArbCom by the community and only modifiable by the community.
• The community soundly rejected in a near-unanimous vote any change to the policy that would permit such actions being taken without notification. No change occurred and arbitrators defended their violation by stating that allowing me to know why I was banned would put the admin at risk of harassment.
• Wikipedia co-founder Jimmy Wales commented months later when I noted he had previously promised to overturn exactly such a proceeding. After confirming it concerned my report about the admin, he argued mine was a special case and that I was only pretending not to know why I was banned.
• Both Wales and members of ArbCom emphatically insisted there was nothing to my report. I investigated further and found more and more evidence validating all of my claims and sent it all to ArbCom. My e-mails to the admin and my reports to the Committee before and after my ban were all posted with redactions.
• During this time I also found evidence members of ArbCom involved in my ban had previously criticized this admin’s editing on some of the exact same articles mentioned in my report.
• My first appeal got my first official response stating they “disagreed” with my claims and falsely suggested I had threatened to reveal the admin’s identity. They offered conditions for lifting my ban, but would not provide clarification on terms when asked.
• After I met their stated conditions for lifting my ban to the best of my ability given lack of clarification, they dropped the claims of harassment and excluded the incident from their reasons for keeping my ban in place.
• The administrator has continued engaging in undisclosed conflict of interest editing this year. I have reason to believe ArbCom is aware of at least some of this editing and that it involves promotional editing violating the site’s content policies and guidelines.