Trump’s social media bias reporting project is a data collection tool in disguise
Three years ago this month, Mark Zuckerberg gathered together a group of influential conservatives to defend Facebook against allegations of political bias. The company had found itself under pressure after Gizmodo reported that the editors who then worked for Facebook “routinely suppressed conservative news” from its since-abandoned Trending Topics module. It hoped that a roundtable discussion with Glenn Beck, Fox News host Dana Perino, and others would quell the growing panic that Silicon Valley liberals were stifling dissent.
Conservatives need not have worried. Nearly every time the analytics firm NewsWhip reports on the top publishers on Facebook, Fox News ranks near the top. (It fell to No. 2 for the first time this year in March, when the Daily Mail edged it out.) But among many conservatives, including many elected officials, it is now an article of faith that social networks discriminate against them.
The past year has seen multiple congressional hearings devoted to trumped-up allegations of bias against social media bias. Today the president — who recently met with Twitter CEO Jack Dorsey to complain that the company’s removal of bots had depressed his follower count — issued a new call for allegations. Makena Kelly has the story:
On Wednesday, the White House launched a new tool for people to use if they feel they’ve been wrongly censored, banned, or suspended on social media platforms like Facebook and Twitter.
“Too many Americans have seen their accounts suspended, banned, or fraudulently reported for unclear ‘violations’ of user policies,” the site reads. “No matter your views, if you suspect political bias caused such an action to be taken against you, share your story with President Trump.”
The “tool” launched by the administration is, in fact, a Typeform page, which can be set up in a few minutes by anyone. The White House’s wording is broad enough that it might inspire anyone who has ever had a bad experience on a social network to register a complaint. “SOCIAL MEDIA PLATFORMS should advance FREEDOM OF SPEECH,” the form shouts. Whether platforms might also seek to moderate hate speech or terrorism (for example) isn’t a question that makes it into the form.
This being the Trump White House, this new tool also appears to be a scam. As Kevin Roose put it: “The thing about the Trump Facebook bias survey is it’s just going to be used to assemble a voter file, which Trump will then pay Facebook millions of dollars to target with ads about how biased Facebook is.”
In the meantime, “bias” is defined ever downward. In conservative parlance, it now refers to any instance in which the user of a social platform did not have a desired outcome. You didn’t appear high enough in search results? Your video wasn’t promoted by an algorithm? You were suspended for threatening to kill someone? It’s all just “bias” now.
And for platforms, that presents a devilish trap. With the definition of “bias” constantly expanding, it becomes harder and harder to argue that it doesn’t happen. Social networks often do make mistakes around content moderation, which they have outsourced to an army of underpaid workers who must deal with near-daily changes to community standards. But the larger truth — that conservatives are thriving in part due to their mastery of social platforms like Facebook and YouTube — forever goes unsaid.
Free speech is more than just a Typeform to the White House. It’s also the rationale by which the United States will break with 18 other governments and refuse to sign the Christchurch Call, a set of voluntary commitments by states and online service providers designed to prevent the dissemination of violent and extremist content.
The White House’s view — to the extent that there is such a thing as a White House apart from the whims of the president — is not coherent. Tony Romm and Drew Harwell report:
The White House felt the document could present constitutional concerns, officials there said, potentially conflicting with the First Amendment., even though Trump previously has threatened to regulate social media out of concern that it’s biased against conservatives. […]
Still, in a statement about the Christchurch call, the White House said it stands “with the international community in condemning terrorist and violent extremist content online,” and supports the call’s goals. But the U.S. is “not currently in a position to join the endorsement.”
The call is a non-binding agreement, so signing it would not place any draconian new limits on speech. But it would require the Trump Administration to commit informally to de-radicalizing citizens, supporting academic research into violent extremism, and collaborating with other countries. And it refused.
Other countries are taking much more aggressive action. Australia and Singapore have proposed onerous new laws against social platforms that require them to remove some content immediately, under penalty of massive fines or even jail time for executives. That helps to explain why the platforms all showed up in Paris today to sign the Christchurch Call, agreeing to take new action to prevent the use of live streaming and other platform technologies to promote terrorist ideologies.
Facebook, which is perhaps under the most scrutiny, went a step further. It introduced new restrictions on live streaming that will prevent people who violate its policies from broadcasting for 30 days. Had the rule been in place, the Christchurch shooter would not have been able to broadcast his attack, the company said.
Social networks have an important role to play in reducing the spread of terrorism. But they need help from the countries in which they operate. It’s heartening that 18 governments today committed to working with them on the project — and beyond dispiriting that the United States, for the most craven of reasons, opted out.
Here’s another consequential order from the president:
After months of speculation, President Trump has signed an executive order giving the federal government the power to block US companies from buying foreign-made telecommunications equipment deemed a national security risk.
Under the order, which gives the secretary of commerce power to determine which transactions may be potential risks, no single company is immediately marked as a threat. But the plan is largely seen as a move against China-based Huawei, which some US lawmakers have deemed a security threat.
Alex Kasprak investigates how evangelical Christians are repurposing Facebook pages and “to build a coordinated, pro-Trump network that spreads hate and conspiracy theories.”
Though the actual authorship of the posts within these pages is opaque, their titles imply diverse representation from a broad swath of American demographic groups, including “Jews & Christians for America” and “Blacks for Trump.” In reality, however, the pages in this network are all connected to evangelical activist Kelly Monroe Kullberg. But she is neither black nor Jewish, and her views appear to represent an extreme subset of the broader evangelical movement in America. Though we do not know for sure what individual or individuals created each of these pages, or if Kullberg, her family members, or various “interns” write their posts, all of them appear now to be tied financially to Kullberg or to organizations she has created. As far as we have been able to ascertain, Facebook has no problem with the existence of this coordinated network, which we will refer to here as the “Kullberg network.”
As a resident, I’m really glad to hear this. From Kate Conger, Richard Fausset and Serge F. Kovaleski:
The action, which came in an 8-to-1 vote by the Board of Supervisors, makes San Francisco the first major American city to block a tool that many police forces are turning to in the search for both small-time criminal suspects and perpetrators of mass carnage.
The authorities used the technology to help identify the suspect in the mass shooting at an Annapolis, Md., newspaper last June. But civil liberty groups have expressed unease about the technology’s potential abuse by government amid fears that it may shove the United States in the direction of an overly oppressive surveillance state.
I’m somewhat pessimistic that we will ever get the majority of people to believe this, but I appreciate CNET fighting the good fight:
In an informal study, CNET reporters discussed predesignated topics in front of their phones and then monitored the devices for related ads. We found nothing to suggest Facebook had overheard our conversations. While we can’t 100% disprove this idea, security experts have also failed to find evidence the social network is eavesdropping on users to target ads more effectively.
And it pays them triple what it currently pays its content moderators in North America. Rani Molla:
Facebook interns earn $8,000 a month — more than any other company’s paid interns — according to a new report by Glassdoor, a site where employees anonymously review their companies and report their salaries. For context, if interns were to work a full year, they would make $96,000, which is nearly double the median pay of $52,807 a year ($4,400 a month) for regular jobs found on the site.
Really sad news, and perhaps a case where machine learning will let Instagram intervene faster over time:
Authorities in Malaysia have launched an investigation after a teenage girl was found to have taken her own life after starting an Instagram poll asking users to vote for life or death.
The unnamed 16-year-old posed the question to users on Monday, saying: “Really important, help me choose: D/L.”
Go off, queen:
“I see these young girls, devastated with bullying and not being able to have my voice.
”It can be great in moments, but I would careful and give yourself time limits in terms of when you can use it and not.”
Nicole Nguyen sits in on a bunch of Twitter meetings where product designers tell us they have very nearly cracked their long-stated mission of rethinking every incentive and interaction on the platform, and then show her a thing where when someone replies to you, you can tap their face and see their profile.
Well that’s unfortunate — and it sure seems like Apple could do something to help here. From Lorenzo Franceschi-Bicchierai:
As of today, there is no specific tool that an iPhone user can download to analyze their phone and figure out if it has been compromised. In 2016, Apple took down an app made by Esser that was specifically designed to detect malicious jailbreaks. Moreover, iOS is so locked down that without hacking or jailbreaking it first, even a talented security researcher can do very little analysis on it. That is why security researchers crave expensive iPhone prototypes that have security features disabled, as a Motherboard investigation revealed earlier this year.
Claudio Guarnieri, a technologist at Amnesty International, who found that a colleague of his was targeted by NSO spyware last year, said that the “irony” is that there are better tools for attackers who want to do forensics on iOS—such as Cellebrite and GrayShift—than for defenders who want to help victims.
Well here’s something novel: Facebook — “a widely used website,” in the description of its employees here — contributing code to Chrome. It’s an effort to make interactions on pages faster:
Charlie Warzel says they’re no true opting out of personal data collection:
We’ve built a society and economy that runs on surveillance, a world where the price for participation is tracking, targeting and disclosure of data. “Opting out” might as well mean heading to Walden Pond (and even then it’s likely that, in preparation for your journey to Thoreau’s cabin, you’d be targeted by ads for self-reliance books on Amazon, freeze-dried prepper meals and 12 different iPhone meditation apps).
And finally …
YTMND, a legendary meme-making site, is apparently no more. I can’t say I visited much in recent years, but during Tumblr’s heyday in the mid-‘00s it was a reliable source of joy. RIP.
Before impact fonts and weird Twitter, there was YTMND.com. The site allowed anyone to attach a gif, usually animated but not always, to a bit of looping sound. Users could vote on these weird animations, remix, and share them. When it launched, it was the perfect mix of ridiculousness and interactivity.
It was named for its first upload—tiled pictures of Sean Connery pointing aggressively while a clip of him saying “you’re the man now dog” from the Gus Van Sant movie Finding Forrester played over and over again on a loop. Back in 2001, we thought this was the height of comedy. Even now, watching the clip brings a smile to my face.
Talk to me
Send me tips, comments, questions, and your thoughts on tech company collaboration: [email protected].
This article is from The Verge