If there’s a worse idea than the Pentagon becoming Editor-in-Chief of America, I can’t remember it. But we’re getting there:
From Bloomberg over Labor Day weekend:
Fake news and social media posts are such a threat to U.S. security that the Defense Department is launching a project to repel “large-scale, automated disinformation attacks,” as the top Republican in Congress blocks efforts to protect the integrity of elections.
One of the Pentagon’s most secretive agencies, the Defense Advanced Research Projects Agency (DARPA), is developing “custom software that can unearth fakes hidden among more than 500,000 stories, photos, video and audio clips.”
Once upon a time, when progressives still reflexively distrusted the military, DARPA was a liberal punchline, known for helping invent the Internet but also for developing lunatic privacy-invading projects like LifeLog, a program to “gather in a single place just about everything an individual says, sees, or does.”
DARPA now is developing a semantic analysis program called “SemaFor” and an image analysis program called “MediFor,” ostensibly designed to prevent the use of fake images or text. The idea would be to develop these technologies to help private Internet providers sift through content.
It’s the latest in a string of stories about new methods of control over information flow that should, but for some reason do not, horrify every working journalist.
From the Senate dragging Internet providers to the Hill to demand strategies against the sowing of “discord,” to tales of hundreds of Facebook sites zapped for “coordinated inauthentic behavior” following advice by government-connected groups like the Atlantic Council, it’s been clear the future of the information landscape is going to involve elaborate new forms of algorithmic regulation.
Stories about the need for such technologies are always couched as responses to the “fake news” problem. Unfortunately, “fake news” is a poorly-defined, amorphous concept that the public has been trained to fear without really understanding.
The term surged into public view three years ago. Experts insisted Macedonian troll farms and pranksters like the late Paul Horner (who once conned Fox News into doing a story that Barack Obama was funding a Muslim culture museum) had an enormous impact on Trump’s victory.
Had they? When “fake news” first became “a thing,” as media critic Adam Johnson put it in The Nation three years ago, I was skeptical.
Fake news has a long history in America. Its most pernicious incarnation is never the work of small-time scam artists. The worst “fake news” almost always involves broad-scale deceptions foisted on the public by official (and often unnamed) sources, in conjunction with oligopolistic media companies, usually in service of rallying the public behind a dubious policy objective like a war or authoritarian crackdown.
From the sinking of the Maine in 1898, to rumors of a union-led socialist insurrection before the Palmer raids in 1919, to the Missile Gap in the late fifties and early sixties (here is the CIA’s own website admitting that one was “erroneous”), to the Gulf of Tonkin lie that launched the Vietnam War, to the more recent WMD fiasco, true “fake news” is a concerted, organized, institutional phenomenon that involves deceptions cooked up at the highest levels.
The other “fake news” – the dubious panic over which began in November-December of 2016 – is a strange, hybrid concept that mixes fear of fever-swamp conservative lunacies with satire, Russian propaganda, legitimate dissent, and other content.
The most infamous example usually cited is Pizzagate, in which Hillary Clinton and campaign chief John Podesta were falsely said to be running a pedophile ring out of the basement of a Washington pizza restaurant. The hoax carried import because a 28 year-old North Carolinian named Edgar Maddison Welch was idiot enough to shoot up the joint in response.
But the other specific examples cited of “fake” news most often cited are patently absurd: that the Pope or the Amish endorsed Donald Trump, that Hillary Clinton sold weapons to ISIS, that an FBI agent investigating Clinton had died in a house fire (a story broken by the nonexistent “Denver Guardian”), that the Democrats paid protesters to heckle Trump events, etc.
The idea that these fake tales had a major impact in 2016 is absurd on its face. They didn’t change things any more than ALIEN BACKS CLINTON swayed the 1992 election.
It was laughable beyond belief to see stories in outlets like the New York Times and the Washington Post taking seriously the notion that small-time hoaxers like Horner — who was trying to sucker Trump fans to websites so he could make maybe ten grand a month off click ads – were a major threat to national security. (That some cited Horner’s own claim of responsibility for Trump’s election was even more preposterous).
When officials calling for a crackdown talk about “fake news,” you’ll often see them conflating examples of provably false stories with true stories circulated or interpreted in undesirable ways: the Clinton email scandal, the Uranium One story, the Podesta email leak, etc.
Even a controversy about Hillary Clinton’s health, cited by Ohio State University researchers as an example of the pernicious impact of fake news, was an amalgam of true and fake.
There was indeed wild speculation on the Internet and by goons like Sean Hannity about Clinton suffering from seizures or dementia. This was mixed in with real events like a 2012 collapse that caused a concussion, the subsequent discovery of a blood clot in Clinton’s brain (ABCnews.com called it “life threatening” in a headline), and Clinton’s September, 2016 collapse at a 9/11 memorial event.
The New York Times, CNN, CBS, the Washington Post and other reputable outlets covered the latter episode in great detail. As Vox noted at the time, this turned an online conspiracy theory into a “mainstream debate.”
If there’s a fake news story out there, it’s the fake news panic itself. It has the hallmarks of an old-school, WMD-style propaganda campaign.
It includes terrifying pronouncements by unnamed “intelligence officials,” unprovable, overblown, or outright fake statistical assertions about the threat (like the oft-cited claim that fake election news had more engagement than real news), open conflation of legitimate domestic dissent with foreign attack, and routine dismissal of experts downplaying the problem (here are two significant studies suggesting the “fake news” phenomenon is overstated).
Of course, the final, omnipresent ingredient in most major propaganda campaigns is the authoritarian solution. Here, it’s unelected, unsupervised algorithmic control over media. We’ve never had a true news regulator in this country, yet the public is being conditioned now to accept one, without thinking of the consequences.
The most enormous issue posed by the modern media landscape is the industry’s incredible concentration, which allows a handful of private platforms – Facebook, Twitter, Google – to dominate media distribution.
This makes it possible to envisage direct levers of control over the public’s media habits that never existed back when people got much of their news from local paper chains with individual distribution networks. We’ve already seen scary examples of misidentified foreign subversion, from the Washington Post’s repeat editorials denouncing Bernie Sanders as a useful idiot for the Kremlin to the zapping of hundreds of domestic political sites as “coordinated inauthentic behavior.”
What if the same people who can’t tell the difference between Truthdig and Pravda get to help design the new fake news algorithms? That’s a much bigger worry than the next Paul Horner or even, frankly, the next Russian Facebook campaign. While Donald Trump is in the White House, progressives won’t grasp how scary all of this is, but bet on it: In a few years, we’ll all wish we paid more attention when the Pentagon announced it wanted in on the news regulation business.