View profile

Senators cry bias, and Facebook crumbles

Revue
 
In 2016, Facebook killed off a part of the service that highlighted trending news articles, following
 
September 11 · Issue #382 · View online
The Interface
In 2016, Facebook killed off a part of the service that highlighted trending news articles, following a hysterical overreaction to a Gizmodo piece that claimed, weakly, that the company was “routinely suppressing conservative news.” (It turns out that when human editors are told to pick the day’s most important stories, they tend to pick reported articles from mainstream sources over hyper-partisan opinion pieces.) Conservative media howled with outrage over the report, Facebook panicked and fired its editors, and the job of serving up links to its user base was outsourced entirely to algorithms, which elevated misinformation above journalism throughout the 2016 presidential election campaign.
Next month, human beings will rejoin the ranks of Facebook editors. The company is working on a new news tab, and humans are going to edit it. Facebook is negotiating to pay publishers what are essentially licensing fees for news content — offering welcome and much-needed direct compensation to organizations that have struggled to compete with the Facebook-Google digital advertising duopoly. The new tab, which seems to be roughly modeled on Apple’s relatively uncontroversial news service, will task editors with picking the day’s most important stories and organizing them. (Algorithms will offer supplemental assistance.)
Unlike previous efforts at Facebook, this time editors will choose stories from a whitelist of publishers rather than simply surface stories that are getting lots of clicks. And, Alex Heath reported this week in The Information, the company hopes to avoid charges of bias by adhering to strict editorial guidelines:
This time around, Facebook hopes to avoid allegations of bias in the news tab by imposing stricter editorial rules for editors and hiring them as full time employees rather than outside contractors, said people familiar with the company’s thinking. Below the top stories selected by editors, the Facebook news tab will show a feed of stories, selected by software algorithms based on the publishers users follow.
Knowing what you know about how charges of bias are levied today, how hard did you laugh at the idea that Facebook’s whitelist of publishers would help the company avoid such charges?
I laughed moderately hard.
As we have discussed a few times around here, “bias” has been defined down to describe any undesired outcome on social media. Did Twitter recommend that you follow a Democrat rather than a Republican? Bias. Did Facebook suspend a conservative activist’s account for breaking one of its rules? Bias. Did a third-party fact checker accurately characterize an anti-abortion post as false?
Well, you can probably imagine what happened next.
Let’s take an uncharacteristically deep dive into a single Facebook post, so that we might better understand what Facebook is up against as it attempts to apply a straightforward set of editorial guidelines to a platform that serves billions of people a day.
Four Republican senators sent a letter to Facebook CEO Mark Zuckerberg this morning, criticizing the social-media platform’s recent “fact check” of pro-life organization Live Action. In a copy of the letter obtained exclusively by National Review, Senators Josh Hawley (Mo.), Ted Cruz (Texas), Kevin Cramer (N.D.), and Mike Braun (Ind.) condemn what they call Facebook’s “pattern of censorship” and call on the group to submit to an external audit.
At the end of last month, Facebook notified Live Action that fact-checkers had given a “false” rating to two videos shared by the group’s president Lila Rose. One featured Rose herself and the other featured Dr. Kendra Kolb, a board-certified neonatologist; both videos included the claim that abortion is not medically necessary. After bestowing a “false” rating on the videos, Facebook prevented Rose and Live Action from promoting or advertising content and alerted users who had shared the two videos that they had spread “false news.”
The Review story omits some key context. Facebook didn’t fact-check Live Action itself; the fact-check was conducted by Science Feedback, a partner with domain-area expertise. You can read the fact-check here. The reviewers’ rationale for labeling the post in question false is straightforward. In the video under review, Rose says “abortion is never medically necessary.” In fact, it sometimes is. The reviewers write:
Physicians who evaluated this claim found it to be inaccurate. They explained that there are many medical conditions, such as pre-eclampsiaHELLP syndrome and placenta previa, in which abortion could become medically necessary in order to save the life of the mother. Incidentally, abortion is medically defined as a procedure to end a pregnancy – this definition does not change depending on the reasons for an abortion, i.e. whether the procedure is motivated by an unwanted pregnancy or medical emergency or some other situation has no effect on its medical definition. However, Lila Rose redefines the meaning of abortion to exclude the cases when abortion is medically necessary in order to bolster her claim that “abortion is never medically necessary”. This is akin to the No True Scotsman fallacy in which the definition of a word/phrase is modified from its actual meaning to make a point. For example, Rose claims that treating an ectopic pregnancy is not an abortion, even though termination of the pregnancy is the result of the procedures that treat ectopic pregnancies.
I trust the physicians on these points; the senators don’t. (The fact that one of the physician reviewers has performed abortions makes the fact-check more credible to me, not less.) But set aside your own beliefs on abortion for the moment, if you can. How can Facebook avoid charges of “bias” when the entire nature of editorial decision-making is to privilege one set of views over another?
It can’t, of course. Charges of bias are here to stay — and I imagine we’ll see many more Congressional hearings on the subject as lawmakers attempt to work the new refs.
Especially because working the refs … works. Here’s what Facebook had to say when I asked about the senators’ complaint:
“Posts by Live Action and Lila Rose were fact-checked by a third party, independently certified by the International Fact Checking Network. We have been in touch with the IFCN which has opened an investigation to determine whether the fact checkers who rated this content did so by following procedures designed to ensure impartiality. While the IFCN investigates, we are removing the relevant fact checks and have communicated this to the members of the US Senate who brought this specific concern to our attention.”
So one letter from Congress later, Rose’s false claim that abortions are never medically necessary is now free to circulate on Facebook until further notice. You can probably imagine what lesson the senators will take away from this.
What happens when a story about abortion that senators dislike appears prominently in the new news tab? Will Facebook respect its editors’ news judgment and back them up? Or will it bow to the sensitivities of lawmakers? I understand the reluctance to let tech platforms shape the boundaries of public discourse. But I’d still rather have journalists deciding which journalism people should read than congressmen.

The Ratio
🔽 Trending down: California lawmakers approved a bill that requires app-based companies like Uber and Lyft to treat contractors as employees. The move, which would allow drivers to seek basic protections like minimum wage and unemployment insurance, has implications for all tech platforms that rely heavily on contract workforces — which is most of them.
Governing
Amazon’s antitrust probe is heating up as the Federal Trade Commission (FTC) interviews merchants to investigate whether the retail giant squashes competition. The merchants, mostly small businesses, are being asked how much money they make on Amazon in comparison to other online retailers like Walmart and EBay. Spencer Soper and Ben Brody report Bloomberg:
The interviews indicate the agency is in the early stages of a sweeping probe to learn how Amazon works, spot practices that break the law and identify markets dominated by the company. The length of the interviews and the manpower devoted to examining Amazon point to a serious inquiry rather than investigators merely responding to complaints and going through the motions, antitrust experts say.
State attorneys general order Google to turn over information about its ad business as part of an ongoing antitrust investigation. They’re seeking intel on the company’s past acquisitions, top advertisers and publishers, pricing models, and data collection and data sharing practices. Here’s Bloomberg reporters David McLaughlin, Gerrit De Vynck, and Naomi Nix:
The process of showing an ad to a single person visiting a web page can involve dozens of companies and multiple auctions and transactions. Google has worked its way into controlling much of that process, and investigators want to know exactly how powerful the company has become in this space.
In addition to scrutiny from antitrust regulators, Google’s acquisition strategy has also drawn the ire of top Democratic lawmakers in Washington, who argue they have enabled the tech behemoth to control far too much of the digital advertising ecosystem, crowding out competitors.
We still have no idea how Russian manipulation will impact the 2020 election, according to an MIT professor. Sinan Aral argues social scientists need to do a better job measuring the impact of social media manipulation. (Patt Morrison / Los Angeles Times)
A fake story alleging that President Trump was donating $1 million to Hurricane Dorian victims in the Bahamas circulated on Facebook. The myth, which started in a QAnon conspiracy theory group, spread amongst Trump supporters before being flagged by Facebook as part of their efforts to combat misinformation. (Daniel Funke / PolitiFact)
The man behind Trump’s 2016 Facebook strategy — Brad Parscale — has talked up his all-American, rags-to-riches origin story. But it seems that he made a lot of it up. (Peter Elkind and Doris Burke / ProPublica)
Fox 29 news anchor Karen Hepp is suing Facebook and Reddit after her photo was used without her consent in advertisements hawking dating apps, erectile dysfunction products, and porn sites. (Victor Fiorillo / Philadelphia Magazine)
Industry
YouTube creators are changing their strategy after a recent FTC settlement left many worried they’d be unable to make money on the platform. As part of the $170 million agreement, YouTube has to stop collecting data on kid-specific content, and creators who target their videos at children below the age of 12 have to clearly label it as such. As Julia Alexander reports at The Verge, creators are adapting to the new rules by creating targeted content for teens, scrapping old series related to toys or games, and even switching to vlogging:
Several prominent full-time personalities are already making changes to avoid the potential fallout. Toya from MyFroggyStuff (2.1 million subscribers), Kelli Maple (1.2 million subscribers), and Rob of Art for Kids (1.9 million subscribers), among others, have all published announcement videos updating their fans about changes they should expect to see, including adjustments to both titles and content. Their goal is to avoid losing revenue by being swept up in YouTube’s broad new category of kids content, which will go into effect on January 1st, 2020, while still keeping their fans happy.
Controversial YouTuber PewDiePie pledged $50,000 to the Anti-Defamation League in an about-face that spurred more conspiracy theories. The YouTuber has previously been accused of anti-semitic behavior. (Makena Kelly / The Verge)
Twitter will now let you rearrange the photos you attached to a tweet before you post it. Previously you had to delete the photos and re-upload them and it was a nightmare.
Instagram is working on a new video feature called “clips,” which looks a lot like TikTok. “Clips” allows users to record short video segments, overlay them with music, and adjust their speed. Sound familiar? (Jane Manchun Wong / wongmjane.com)
A travel influencer who admitted to faking clouds in her Instagram photos just got a job with a photo editing app, Enlight Quickshot, to produce even more fake cloud patterns. Success! (Tanya Chen / BuzzFeed)
Recode just launched a new podcast, Reset, about how tech is changing our lives. It’s hosted by Arielle Duhaime-Ross and launches on October 15. Subscribe! (Liz Nelson / Vox)
And finally ...
Twitter Suspends an Account for Tweeting a Cartoon of Captain America Punching a Nazi
Talk to us
Send us tips, comments, questions, and your guidelines for Facebook news editors: casey@theverge.com and zoe@theverge.com.
Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here
If you were forwarded this newsletter and you like it, you can subscribe here
Powered by Revue