On Wednesday, Rob Price published a nice scoop about a San Francisco marketing agency that had been slurping up millions of public posts from Instagram and selling it to clients. As a side effect, the company — which is called HYP3R, because San Francisco — took countless Instagram stories that were supposed to disappear after 24 hours and effectively made them permanent, storing them in a database and then renting it out to brands.
The FBI wants to build a data dragnet on Facebook
And it seems to go against Facebook’s privacy policies
And it seems to go against Facebook’s privacy policies
Illustration by William Joel / The Verge
is the editor of the Platformer newsletter and cohost of the Hard Fork podcast.
All of this is against Instagram’s terms of service, though the company didn’t notice it until Price brought it to their attention. I’m told that the practice was not particularly uncommon — one tipster told me yesterday that many marketing tech companies sell services that let brands see what people are saying about them on Instagram and other social networks. Scraping the web for commercial purposes is as old as the web itself — but as more Americans grow more concerned about privacy issues, practices like HYP3R’s are making us uncomfortable.
And as it turns out, marketing tech companies aren’t the only organizations who want to scrape social data for their own purposes. For example, there’s the Federal Bureau of Investigations. Here are Jeff Horwitz and Dustin Volz in the Wall Street Journal today:
The FBI is soliciting proposals from outside vendors for a contract to pull vast quantities of public data from Facebook, Twitter and other social media “to proactively identify and reactively monitor threats to the United States and its interests.” The request was posted last month, weeks before a series of mass murders shook the country and led President Trump to call for social-media platforms to do more to detect potential shooters before they act. The deadline for bids is Aug. 27.
As described in the solicitation, it appears that the service would violate Facebook’s ban against the use of its data for surveillance purposes, according to the company’s user agreements and people familiar with how it seeks to enforce them.
News of the FBI’s interest in Facebook comes in the same week that the president called on social networks to build tools for identifying potential mass murderers before they act. And across the government, there appears to be growing consensus that social networks should become partners in surveillance with the government.
But so far, as the Journal story illustrates, the government’s approach has been incoherent. On one hand, it fines Facebook $5 billion for violating users’ privacy; on the other, it outlines a plan to potentially store all Americans’ public posts in a database for monitoring purposes.
It seems like we should probably only do one of those things.
What should social networks’ role in law enforcement? It would be wonderful if Facebook, Twitter, or Reddit could reliably identify murderous manifestoes as they are posted and alert the authorities. But as my colleague Adi Robertson has noted, such a system would almost certainly generate many false positives — giving conservatives in particular yet more reason to complain that they are being censored by liberal Silicon Valley elitists.
And while I assume our national security apparatus already pays close attention to social networks, I still recoil at the idea of it building a comprehensive database of public posts.
Anyway, isn’t there maybe something else the government could do here? Like, say, make combating domestic terrorism a priority at the Department of Homeland Security?
White House officials rebuffed efforts by their colleagues at the Department of Homeland Security for more than a year to make combating domestic terror threats, such as those from white supremacists, a greater priority as specifically spelled out in the National Counterterrorism Strategy, current and former senior administration officials as well as other sources close to the Trump administration tell CNN.
”Homeland Security officials battled the White House for more than a year to get them to focus more on domestic terrorism,” one senior source close to the Trump administration tells CNN. “The White House wanted to focus only on the jihadist threat which, while serious, ignored the reality that racial supremacist violence was rising fast here at home. They had major ideological blinders on.”
I’m sure that social networks have some role to play in fighting terrorism that goes beyond their current efforts. But that’s even more true of the federal government, which is still catching up to the nature of the threat. Perhaps we should start there.
Democracy
Well this seems bad. From Kim Zetter:
For years, U.S. election officials and voting machine vendors have insisted that critical election systems are never connected to the internet and therefore can’t be hacked.
But a group of election security experts have found what they believe to be nearly three dozen backend election systems in 10 states connected to the internet over the last year, including some in critical swing states. These include systems in nine Wisconsin counties, in four Michigan counties, and in seven Florida counties—all states that are perennial battlegrounds in presidential elections.
Jonathan Stempel reports on what could be a highly consequential decision:
A federal appeals court on Thursday rejected Facebook Inc’s effort to undo a class action lawsuit claiming that it illegally collected and stored biometric data for millions of users without their consent.
The 3-0 decision from the 9th U.S. Circuit Court of Appeals in San Francisco exposes Facebook to billions of dollars in potential damages to the Illinois users who brought the case.
I absolutely dread having to cover whatever accursed hearing comes out of this one:
President Donald Trump’s campaign and the major, national Republican campaign committees announced Thursday they will stop spending money to advertise on Twitter after the social media platform locked Majority Leader Mitch McConnell’s account for posting a video of violent threats against him. […]
The moves comes after McConnell’s campaign account was locked Wednesday for posting a video of protesters outside his home that included violent threats against the Kentucky Republican. A Twitter spokesperson confirmed the account was temporarily locked for violating the company’s “violent threats policy.” McConnell’s campaign manager, Kevin Golden, said the campaign unsuccessfully appealed the decision and accused Twitter of hypocrisy.
“About one in four companies revealed personal information to a woman’s partner, who had made a bogus demand for the data by citing an EU privacy law,” Leo Kelion reports:
The security expert contacted dozens of UK and US-based firms to test how they would handle a “right of access” request made in someone else’s name.
In each case, he asked for all the data that they held on his fiancee. In one case, the response included the results of a criminal activity check. Other replies included credit card information, travel details, account logins and passwords, and the target’s full US social security number.
Shannon C. McGregor and Daniel Kreiss hunt for reasons why claims of “censorship” are growing. The story only focuses on claims related to advertising, but it’s still worth reading:
Facebook and Google seldom disclose much about how they make decisions on moderating political content. Both firms require campaigns to adhere to a broad set of advertising standards that can be interpreted very flexibly. For example, Google bans “inappropriate content” such as “intimidation” and “discrimination,” but it says nothing about what these things mean in practice.
Here’s one example. Political practitioners told us that advertising that focuses on the politics of assault rifles, for or against, might run afoul of Google’s rules. At Google, algorithms vet most advertising for “inappropriate content.” When an algorithm flags an ad, it then goes to human reviewers. If reviewers reject the ad, they give very little explanation — failing to clarify, for instance, why an ad about the politics of assault rifles counts as “inappropriate content.” As a result, campaigns don’t know how to design ads that meet the standards; that limits the range of political topics on which politicians can campaign.
Anna Wiener has a beautiful profile of Daniel Gackle and Scott Bell, content moderators for Hacker News, a popular online tech-focussed forum owned by the startup accelerator Y Combinator.
On Facebook and YouTube, moderation is often done reactively and anonymously, by teams of overworked contractors; on Reddit, teams of employees purge whole user communities like surgeons removing tumors. Gackle and Bell, by contrast, practice a personal, focussed, and slow approach to moderation, which they see as a conversational act. They treat their community like an encounter group or Esalen workshop; often, they correspond with individual Hacker News readers over e-mail, coaching and encouraging them in long, heartfelt exchanges. […]
He and Bell assert their own opinions in subtle ways. Recently, they made some small changes to the Hacker News guidelines, which have always hewed closely to those that Graham drafted in 2007. To one about throwaway accounts—acceptable for sensitive information but discouraged as a regular practice—they added the reminder “HN is a community.” In another—“Comments should get more civil and substantive, not less, as a topic becomes more divisive”—they changed the phrase “civil and substantive” to “thoughtful and substantive.”
Lachlan Markay reports on “an effort to implicate Bolton in a global drug trafficking ring,” saying it “bears hallmarks of past Iranian influence operations.”
On Monday, a Twitter user claiming to be a high-ranking Canadian law enforcement official posted records supposedly showing a $350,000 wire transfer from a Canadian children’s apparel company to a Swiss bank account owned by National Security Adviser John Bolton’s daughter. “Police investigations show [the company] and its CEO are accused of laundering and transferring dirty money between Canada and some European countries, Turkey, the United Arab Emirates and the United States,” the Twitter account claimed.
The claims are clearly fabricated, and the effort does not appear particularly sophisticated. But a U.S. official familiar with the apparent disinformation campaign said intelligence community officials were aware of the effort. And Lee Foster, an information operations intelligence analyst at the cybersecurity firm FireEye, told The Daily Beast that the hoax’s techniques are “consistent with what we’ve seen with previous pro-Iranian influence operations.”
“Domestic terrorism” isn’t recognized as a crime the way international terrorism is, Jane Coaston reports. It should be:
For many reasons, domestic terrorism in the United States — largely committed by people ideologically tied to white nationalism and white supremacy — is viewed differently from the international variety and has no specific penalties under the law. This fact has been the source of anger on the left for years — after the Oklahoma City bombings, for example — but some prominent conservatives have now coalesced around the same thought.
Even President Donald Trump described the El Paso shootings as terror, saying, “We have asked the FBI to identify all further resources they need to investigate and disrupt hate crimes and domestic terrorism, whatever they need.” And commentators across the political spectrum seem more willing than ever to view white nationalist terrorism as what it is.
Elsewhere
Is my dream of Facebook paying the equivalent of carriage fees to high-quality publishers actually on the horizon? Benjamin Mullin and Sahil Patel report that it just might be:
Facebook is offering news outlets millions of dollars for the rights to put their content in a news section that the company hopes to launch later this year, according to people familiar with the matter.
Representatives from Facebook have told news executives they would be willing to pay as much as $3 million a year to license entire stories, headlines and previews of articles from news outlets, the people said.
In the words of Matt Stoller: “How much of a monopoly do you have to be to literally not care if your cash register breaks? Oh I can’t take your money nvm I’ll get it later.” Kurt Wagner:
But in the past five months, the “Facebook & Instagram Pro Ad Buyers Industry Group” has taken on another purpose: Emotional support. Herrmann’s group has become a place to commiserate about the state of Facebook Ads Manager, the software system used to run advertising campaigns on Facebook and Instagram. Ads Manager is crashing with regularity, according to interviews with numerous advertisers. The outages, which can last for hours at a time and make it impossible to start a new ad campaign or manage an existing one, seem to happen every month, they say.
With almost 900 members, about 170 of whom joined in the last 30 days, Herrmann’s private group has grown quickly alongside Facebook Inc.’s technical stumbles. Screenshots from the group show frustrated advertisers seeking answers about error messages they receive, or sharing updates on bugs they’re waiting on Facebook to fix.
In addition to the ad server going down all the time, media buyers also find Facebook reps arrogant, Kerry Flynn reports:
Media buyers say they’re becoming increasingly frustrated with ads reps from Facebook and Google, even as Twitter’s and Snap’s teams have matured and grown to be more thoughtful and proactive partners.
Facebook’s and Google’s ad reps have a reputation for being slow to respond and even when they eventually reply providing worthless advice, buyers said. Meanwhile, Twitter’s and Snap’s teams have matured over the last year as they refocused their messaging and prioritized communicating them with buyers, no matter the size of their budgets. Buyers say this difference hasn’t yet affected where they spend their ad dollars — Facebook still wins for ROI for many advertisers — but it’s made their jobs more difficult. On the platforms’ part, reps at Twitter and Snap are openly touting their closer relationships with buyers.
The gradual transformation of Instagram into merely an alternate UI for Facebook continues, Kurt Wagner and Sarah Frier report:
Facebook Inc. is taking the first major step in a plan to merge its systems and let users exchange messages among all its different mobile apps – and is chipping away at the independence of Instagram’s direct-messaging product in the process.
Engineers are working to rebuild Instagram’s chat feature using Facebook Messenger’s technology, according to people familiar with the matter. That will make it possible for Instagram users to communicate with those using Messenger, something they can’t do now. To make that technically easier, Instagram’s direct-messaging staff now reports to the Facebook Messenger team, said the people. The look of the photo-sharing app’s messaging product, called Instagram Direct, won’t change much, but the underlying technology powering the service will, the people said.
Natasha Lomas reports on more privacy problems at Twitter:
Twitter has disclosed more bugs related to how it uses personal data for ad targeting that means it may have shared users data with advertising partners even when a user had expressly told it not to.
Back in May the social network disclosed a bug that in certain conditions resulted in an account’s location data being shared with a Twitter ad partner, during real-time bidding (RTB) auctions. In a blog post on its Help Center about the latest “issues” Twitter says it “recently” found, it admits to finding two problems with users’ ad settings choices that mean they “may not have worked as intended”.
Oh my:
The video shows Houts, who has more than 330,000 subscribers, recording a video, and seemingly smacking and spitting on her dog intermittently. The unedited version of the video was uploaded to YouTube as a mistake and was taken down shortly after. YouTube did not respond to the The Verge about whether further action is being taken against Houts’ channel at the time of writing. The site prohibits videos that include or incite abuse toward animals.
Launches
Sure, why not:
Facebook wants its users to pay it directly for the subscriptions, which they can also buy plenty of other places. Facebook and its partners hope they can convert a portion of the platform’s huge audience, which consumes video and other content for free, into paying customers.
Facebook won’t say how much it is charging video companies to sell their stuff. But Larry Fitzgibbon, the CEO of Tastemade, says Facebook is providing “similar economics” to other platforms that sell its Tastemade Plus service, like Apple. So assume Facebook is keeping around 30 percent of each monthly subscription it sells.
Takes
Noam Cohen says forget about video games and pay more attention to how tech platforms manipulate our, uh, attention:
THE MISPLACED FOCUS on violence leaking from pretend digital worlds to the real world misses the glaring digital influence we’re subjected to every day: Silicon Valley companies manipulating users, trying to shape their behavior online and off to make money. That manipulation requires careful planning and testing. Notifications and rewards; tapping into anger, fear, and greed. Something as ill-defined as shoot-gun-online, shoot-gun-offline won’t get the job done, researchers have concluded.
This manipulation is sometimes known as “gamification”—making the mundane more exciting by keeping score, awarding prizes, raising the stakes. The makers of Grand Theft Auto V recently internalized that strategy with an audacious plan to have an operational casino within its online extension, including slot machines, horse racing, blackjack, and poker.
And finally ...
Brands do many strange and surprising things on YouTube, but this one really just hit me out of nowhere:
The company that delights in fetishizing physical objects just got a whole lot more, well, intimate. That’s right, on Aug. 7 Apple dropped a host of ASMR videos designed to excite and titillate the internet’s senses, and the entire thing is as baffling as you’d imagine.
Described as “Apple ASMR: Season 1” — confoundingly suggesting a second season may be in the works — the four videos invite the Apple devotee to, for example, “Unwind with the whispered legend of Ghost Forest.”
Bye!
Talk to me
Send me tips, comments, questions, and legends of Ghost Forest: [email protected].
Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.
