Hate speech is finding a home on InstagramOctober 31, 2018
The aftermath of a national emergency follows a now-familiar pattern. Various bad actors race to fill social platforms with misinformation and outright hate speech. Reporters perform simple searches for conspiracy theories and offensive keywords, and write up stories documenting what they find. Platforms belatedly issue contrite statements, saying there is no place for this kind of thing on their platforms, all evidence to the contrary.
Lather, rinse, retweet.
In the wake of last week’s domestic terror attacks, though, a new vector for bile has emerged. More than ever before, journalists are finding vast swaths of hate speech on Instagram. Yesterday, I mentioned this story from the New York Times that found nearly 12,000 posts with the hashtag “#jewsdid911,” suggesting Jews are responsible for the events of September 11th.
Today, other outlets dived in.
In The Daily Beast, Will Sommer examines how right-wing personalities who were considered too noxious for even Twitter have been granted refuge on Instagram, including Alex Jones, Proud Boys founder Gavin McInnes, the right-wing comedian Owen Benjamin, and Milo Yiannopoulos.
Yiannopoulos caused a stir last week when he lamented in an Instagram post that the mail bombs sent to Democratic leaders did not go off — a post that the company initially decided to leave up before reversing course after a public outcry.
Meanwhile, Taylor Lorenz spoke with Kate Friedman Siegel, who found Instagram unresponsive when she reported anti-Semitic direct messages last month. Siegel, whose handle is “crazyjewishmom,” has more than 800,000 flowers.
Siegel shared screenshots of two anti-Semitic memes she had received via Instagram direct message in September, both of which the platform failed to take action on. One featured an oven with the phrase Jewish stroller plastered on top. The other was a Dr. Seuss parody book cover titled “Horton Hears a Jew,” by “Dr. Goebbels.” […]
Siegel has received anti-Semitic messages since she started the account. But recently, they’ve reached a fever pitch. Siegel said she’s been tagged in memes that depict Adolf Hitler doing the Nazi salute, people desecrating the Israeli flag, people Photoshopping her as Anne Frank, people joking about putting her into a gas chamber, and worse. She reports the ones she sees to Instagram, but she gets so many notifications that there’s no easy way for her to keep track of which reports Instagram has taken action on and which it hasn’t.
It’s tempting to note here that the apparent surge in harassment comes just weeks after Instagram’s cofounders quit, marking the end of the service’s pseudo-independence. But the mechanics that enable harassment and hate speech long predate their exit. I wrote last week about some of the ways Instagram enables bigots to self-organize:
The right wing adopted the hashtag #Soros to share many of these memes, and Instagram helpfully organized the most-engaged posts algorithmically. It auto-populated suggested searches for anyone who began to search for Soros: “soros caravan,” “soros bomb,” “soros jew,” all of which could lead users to further misinformation.
Instagram search results also auto-populated with a bunch of obviously fake Soros accounts, although many of them appear to have been taken down overnight.
That’s at least three big problems for Adam Mosseri and his team to consider: should we provide platforms for people that Facebook has banned, and why? Are we making adequate investments in content moderation? And how do the unique mechanics of our platform enable bad actors to organize and spread hate speech?
As recently as August, Instagram was written about as an oasis — the last refuge for people interested in some good old fashioned social networking. Events of the past several weeks suggest that time has come to an end. We should now expect hate speech to proliferate on Instagram just as everywhere else.
And when the next calamity arrives, Instagram can bet that journalists will be looking.
William Turton did a stunt. Facebook has said it’s fixing the “paid for” field, which lets verified advertisers write whatever they want in a free-form box. But it’s looking like the election will be over then:
On the eve of the 2018 midterm elections, a VICE News investigation found the “Paid for by” feature is easily manipulated and appears to allow anyone to lie about who is paying for a political ad, or to pose as someone paying for the ad.
To test it, VICE News applied to buy fake ads on behalf of all 100 sitting U.S. senators, including ads “Paid for by” by Mitch McConnell and Chuck Schumer. Facebook’s approvals were bipartisan: All 100 sailed through the system, indicating that just about anyone can buy an ad identified as “Paid for by” by a major U.S. politician.
Alexis Madrigal finds a reversal from the 2016 trend:
According to an Atlantic analysis of the top 100 spenders, left-leaning candidates and causes spent $9.43 million from October 21 to October 27 alone, the most recent period for which Facebook data is available. The big spenders on the right spent only $2.65 million.
The top 100 ad buyers on Facebook included 55 left-leaning organizations and candidates, compared to just 22 from the right wing. The remaining 23 big spenders focused on state propositions or corporate campaigns.
Ezra Klein explores why the president needs to make the media his enemy — and how the press can’t resist playing along:
Trump, in other words, manipulates the media using the same tactics as a run-of-the-mill alt-right troll, and for much the same reason: He wants the media to fight with him so he gets more coverage and shows how biased they are against him. He wants the media to fight him because that drives attention to the things he’s saying, to the conspiracies he’s popularizing, and to himself. Going to war with the media nets Trump much more coverage than giving a speech on manufacturing policy or tax cuts.
The problem is Donald Trump isn’t your run-of-the-mill troll. He’s the president of the United States of America.
Twitter put up a midterm elections portal to showcase tweets about politics, and naturally it was flooded with fake claims and posts from trolls. Everything about this has me alternating from facepalm to headdesk. Charlie Warzel and Ryan Mac:
A spokesperson for Twitter told BuzzFeed News that the midterms elections portal was like many of the pages the company creates for events like basketball games and natural disasters. An algorithm curates tweets based on certain keywords, and any content associated with those keywords is populated on the page. The Twitter spokesperson did not specify what keywords were being used for its midterms elections page. In a follow-up call, a Twitter spokesperson clarified that the ‘latest’ tab’s algorithm takes into consideration the accounts that a user follows as one of a number of signals in order to surface tweets. It is possible then, if a user follows a conspiratorial or automated feed, the algorithm could recommend similar conspiratorial or automated accounts. The page, the company noted, was designed less for Twitter power-users and more for newer users searching for useful news ahead of the election.
Speaking of Twitter-related facepalms:
Late last week, about 60 percent of the conversation was driven by likely bots. Over the weekend, even as the conversation about the caravan was overshadowed by more recent tragedies, bots were still driving nearly 40 percent of the caravan conversation on Twitter. That’s according to an assessment by Robhat Labs, a startup founded by two UC Berkeley students that builds tools to detect bots online.
Ahead of a visit to India, Chris Daniels, who took over running WhatsApp earlier this year, gave an email interview to Shelley Singh. Daniels mostly delivers talking points, and it’s remarkable how hard he works to shift the conversation toward WhatsApp’s potential as a business tool. Here’s a cute anecdote straight out of the Sheryl Sandberg earnings-call playbook:
A key focus of my visit is how WhatsApp can support Indian businesses and drive economic growth. I believe there are over 50 million small businesses in India. WhatsApp is building tools to help these businesses connect with their customers, respond to questions, and close sales.
An eyewear company in Bengaluru called Glassic told us that 30% of their new sales come from WhatsApp chats. We’re announcing this week a project with Startup India where WhatsApp will directly support new Indian-owned businesses and entrepreneurs.
Facebook earnings were a mixed bag. User growth is slowing — but the company is still growing, despite everything. More interesting was the investor call, where Mark Zuckerberg said stories are poised to overtake feed sharing someday, and that Facebook Watch was becoming a faster-growing destination for watching video.
Microsoft may be getting a new neighbor:
Redmond, Washington, has been the home of Microsoft for decades, but the maker of Word may soon have company. Facebook is planning a 650,000-square-foot office project in the city, according to a report Tuesday.
Ina Fried looks at a report from a coalition of civil rights group calling on social platforms to do more to control the spread of hate speech. As she notes, platforms’ enforcement capabilities are probably more important than the policies themselves, which largely already prohibit these activities.
Specifically, the coalition recommends sites prohibit “hateful activity,” which it defines as “activities that incite or engage in violence, intimidation, harassment, threats, or defamation targeting an individual or group based on their actual or perceived race, color, religion, national origin, ethnicity, immigration status, gender, gender identity, sexual orientation, or disability.”
Twitter released some data on what happened when it expanded its character limit to 280:
To start with, there’s been a 54 percent increase in the use of “please” and a 22 percent increase in the use of “thank you,” because there’s now room for more words. Twitter’s next point will make grammar purists happy: it says it’s seeing a decline in the use of abbreviations like “gr8” (-36%), “b4” (-13%) and “sry” (-5%). More users now tend to type the whole words, since they don’t have to worry about the character limit that much.
Vox.com’s Carlos Maza has a valuable look at “asymmetric polarization” — what happened when Republicans began breaking democratic norms, exposing media outlets’ commitment to bothsidesism.
This is absolutely perfect place to put James Patterson’s next novel and I won’t hear otherwise.
Kara Swisher is not impressed with social media companies’ response to the recent acts of domestic terrorism:
Social media platforms — and Facebook and Twitter are as guilty of this as Gab is — are designed so that the awful travels twice as fast as the good. And they are operating with sloppy disregard of the consequences of that awful speech, leading to disasters that they then have to clean up after.
And they are doing a very bad job of that, too, because they are unwilling to pay the price to make needed fixes. Why? because draining the cesspool would mean losing users, and that would hurt the bottom line. Consider this: On Monday, New York Times reporters easily found almost 12,000 anti-Semitic messages that had been uploaded to Instagram in the wake of the synagogue attack.
And finally …
Here’s an Instagram ghoul that isn’t from the alt-right: Maya Hwang, a young girl from the Philippines who trick-or-treated in a costume that made it appear as if her head were on a plate. This video, in which a neighbor puts candy in Maya’s head-hole, is exactly the sort of absurdity I needed today.
Talk to me
Send me tips, comments, questions, and honestly probably just cute photos of your kids in their Halloween costumes, it would bring me joy during this bleak weak: email@example.com.