Content Moderation Case Study: Facebook Knew About Deceptive Advertising Practices By A Group That Was Later Banned For Operating A Troll Farm (2018-2020)
Furnished content.
Summary:In the lead-up to the 2018 midterm elections in the United States, progressive voters in seven competitive races in the Midwest were targeted with a series of Facebook ads urging them to vote for Green Party candidates. The ads, which came from a group called America Progress Now, included images of and quotes from prominent progressive Democrats including Bernie Sanders and Alexandria Ocasio-Cortez with the implication that these politicians supported voting for third parties. The campaign raised eyebrows for a variety of reasons: two of the featured candidates stated that they did not approve the ads, nor did they say or write the supposed quotes that were run alongside their photos, and six of the candidates stated that they had no connection with the group. The office of Senator Sanders asked Facebook to remove the campaign, calling it “clearly a malicious attempt to deceive voters.” Most notably, an investigation by ProPublica and VICE News revealed that America Progress Now was not registered with the Federal Election Commission nor was any such organization present at the address listed on its Facebook page.In response to Senator Sanders’ office, and in a further statement to ProPublica and VICE, Facebook stated that it had investigated the group and found no violation of its advertising policies or community standards.Two years later, during the lead-up to the 2020 presidential election, an investigation by the Washington Post revealed a “troll farm”-type operation directed by Rally Forge, a digital marketing firm with connections to Turning Point Action (an affiliate of the conservative youth group Turning Point USA), in which multiple teenagers were recruited and directed to post pro-Trump comments using false identities on both Facebook and Twitter. This revelation resulted in multiple accounts being removed by both companies, and Rally Forge was permanently banned from Facebook.As it turned out, these two apparently separate incidents were in fact closely connected: an investigation by The Guardian in June of 2021, aided in part by Facebook whistleblower Sophie Zhang, discovered that Rally Forge had been behind the America Progress Now ads in 2018. Moreover, Facebook had been aware of the source of the ads and their deceptive nature, and of Rally Forge’s connection to Turning Point, when it determined that the ads did not violate its policies. The company did not disclose these findings at the time. Internal Facebook documents, seen by The Guardian, recorded concerns raised by a member of Facebook’s civic integrity team, noting that the ads were “very inauthentic” and “very sketchy.” In the Guardian article, Zhang asserted that “the fact that Rally Forge later went on to conduct coordinated inauthentic behavior with troll farms reminiscent of Russia should be taken as an indication that Facebook’s leniency led to more risk-taking behavior.”Company considerations:- What is the best way to address political ads that are known to be intentionally deceptive but do not violate specific advertising policies?
- What disclosure policies should be in place for internal investigations that reveal the questionable provenance of apparently deceptive political ad campaigns?
- When a group is known to have engaged in deceptive practices that do not violate policy, what additional measures should be taken to monitor the group in case future actions involve escalations of deceptive and manipulative tactics?
Issue considerations:- How important should the source and intent of political ads be when determining whether or not they should be allowed to remain on a platform, as compared to the content of the ads themselves?
- At what point should apparent connections between a group that violates platform policies and a group that did not directly engage in the prohibited activity result in enforcement actions against the latter group?
Resolution:A Facebook spokesperson told The Guardian that the company had “strengthened our policies related to election interference and political ad transparency” in the time since the 2018 investigation, which revealed no violations by America Progress Now. The company also introduced a new policy aiming to increase transparency regarding the operators of networks of Facebook Pages.Rally Forge and one of its page administrators remain permanently banned from Facebook following the 2020 troll farm investigation, while Turning Point USA and Turning Point Action deny any involvement in the specifics of either campaign, and Facebook has taken no direct enforcement action against those groups.Originally posted to the Trust and Safety Foundation website.
Read more here
posted at: 12:00am on 13-Jan-2022
path: /Policy | permalink | edit (requires password)
The World Handled A 'Wordle' Ripoff Just Fine Without Any IP Action
Furnished content.
In the video game space, it has become commonplace to see creators freak out over "rip-offs" and "clones" of their games when the targets of their ire are actually not rip-offs or clones at all. This typically comes down to the all to common confusion over whether you can own or protect ideas versus specific expression. Typically in these stories, it turns out someone is complaining that they're seeing a similar idea in other games, whether it's first person shooters that share common features, the explosion of battle royale games, or even just artwork.Which brings us to Wordle, a browser-based game that I gleefully enjoy telling my fellow Techdirt readers I have not played. However you feel about the game, it's notable in that its creator has been adamant about not monetizing the game, nor has he bothered registering any copyright or trademark for it. Between that and the game's popularity, there is a ton of goodwill there, which may explain why the world smacked down another person's attempt to actually clone (basically) the game into a mobile app that then required paid subscriptions for all of the features.“I love Wordle so much I decided to make my own Wordle app but with a twist!” tech entrepreneur Zack Shakked wrote on Twitter yesterday. “There’s not just 5-letter words, but also 4, 6, and 7 letter words too! You can also play unlimited times if you’re on the Pro version.”The new version of Wordle on the App Store didn’t just have bigger puzzles, it also required you to pay a subscription to unlock all of its features. It was a greedy innovation that, in the words of Any Baio, compounded the plagiarism into a “naked cash grab.” The downloads, reviews, and active paid trials instantly started rolling in. Yes, they rolled in. And Shakked took to Twitter to brag about the adoption rate in the most irritating way possible."We're going to the fucking moon." That prediction turned out to be true, by which I mean the creator ended up in a place where all the oxygen had been sucked into a vacuum. See, that goodwill I mentioned before for the original Wordle creator led to the internet losing its damned fool mind over this cloning of the Wordle game into a money-grabbing app. In addition to Andy Baio dunking on Shakked, so did many, many others.Cabel Sasser of indie studio Panic, the publisher behind Untitled Goose Game currently working on the Playdate handheld, responded to Shakked that he couldn’t wait to show Wordle’s true creator how to navigate the App Store takedown process. Others were more explicit. “Absolutely fuck you,” wrote Vlambeer cofounder Rami Ismail. And eventually, Apple took the Shakked's app down. The reasons why are trivially easy to understand: the app called itself "Wordle", had nearly the exact same gameplay and look as the original game, and was clearly attempting to profit off of the work of others.But notably absent in all of this was anything related to actual intellectual property registrations and the like. The world managed to take the right and corrective action on a bad actor without talking about copyright, trademarks, lawyers, cease and desist notices, DMCA takedowns, or any of that. You can chalk some of this up to our general ownership culture I suppose, but the truth is this all feels like the public doing its thing via a backlash on someone doing something shitty and Apple responding to that.Which leads to the obvious question: why can't this be the norm?
Read more here
posted at: 12:00am on 13-Jan-2022
path: /Policy | permalink | edit (requires password)
|
|
