Tag: google

Senate Hearings:  A Sea Change for Social Media Companies & Users

Yesterday afternoon the Senate Judiciary Committee held a hearing entitled:  “Extremist Content and Russian Disinformation Online:  Working with Tech to Find Solutions.” Representing the social media companies were Colin Stretch, General Counsel at Facebook; Sean Edgett, Acting General Counsel at Twitter; and Richard Salgado, Director of Law Enforcement And Information Security at Google.

The news to come out of this hearing will not compete with the blockbuster revelations produced the day before by Special Prosecutor Robert Mueller; but in the long run, it may prove to be more important.  Because regardless of who in the current administration may yet be implicated in Russia’s disinformation campaign aimed at the United States, the matters of greatest significance are that it happened, the ways in which it happened, and that is still happening.  And it’s not all about Russia.

Some Background

Shorthand terms like “Russian hacking” do not properly describe the nature of what’s going on; and the significance of what’s going on should be understood as separate from any collusion that may or may not have existed between Russia’s agents and the Trump campaign.  In a nutshell, what the Russian-based Internet Research Agency engaged in had less to do with backing a particular candidate and far more to do with spreading mass dis-information to exacerbate divisiveness among the American electorate.  And there is no better way to achieve this disruption than by using social media platforms.

The estimates reported state that 126 million Americans were exposed to paid, targeted messaging used to spread false and emotionally-charged rumors, some of which favored one candidate or another, but all of which was designed to foment political discord and volatility. As the opening testimony of Clint Watts of the Foreign Policy Research Institute stated in Part II of these Committee hearings, “Terrorists’ social media use has been acute and violent, but now authoritarians have taken it to the next level using social media more subtly to do something far more dangerous – destroy our democracy from the inside out through information campaigns designed to pit Americans against each other.”

That theme—pitting Americans against each other—cannot be overstated in this story, and I’ll return to it shortly.

A Taste of the Hearing

Coming to terms with the negative effects the “information age” can have on democracy is a reckoning long overdue, and yesterday was the first time in my experience that representatives of Silicon Valley were compelled to stifle their utopian rhetoric and admit that their products yield unintended and poisonous consequences.  In fact, early in the hearing, Senator Sheldon Whitehouse (D-RI) directly asked the three witnesses if they were going drop the “we’re just a neutral platform” posturing and accept that they have an active role to play in addressing the matters before the Committee.  All three answered the senator in the affirmative.

That in itself is big news.  The Committee’s unwillingness to accept the shrug of “neutrality” from these companies has implications for cyberlaw that go beyond addressing the immediate issue of foreign powers meddling in US elections.  For instance, it is worth remembering that while Mr. Salgado was promising that Google will take affirmative action and not hide behind a veil of neutrality regarding issues addressed in this hearing, parent company Alphabet’s juggernaut of lobbyists and PR outlets are presently trying to kill the anti-sex-trafficking bill SESTA on the grounds that it would weaken the neutral position of their platform.

There were a few awkward moments between the Committee and the witnesses regarding broader questions about the capabilities of the platforms.  Those of us who advocate certain legal boundaries online (like copyright enforcement) are used to the shell game in which the platforms boast about their capabilities to advertisers one moment (e.g. the ability to perform granular-level, targeted marketing) and in the next moment, state contradictorily that they cannot weed out toxic or illegal content because they “can’t police the internet.”

Among the highlights on this theme was Senator Al Franken’s (D-MN) entertaining inquiry directed at Facebook’s Mr. Stretch in which he asked how, with the company’s extraordinary computing capacity, it failed to “connect two dots” and consider that “American” political ads paid for with rubles might be a reason to doubt the nature of the advertiser.  In a related exchange with Mr. Salgado on the subject of Google’s capacity to weed out foreign-based political ads, Sen. Franken felt the response was too internal-policy focused and reminded the witness, “You know it’s illegal for any foreign money to be spent in our election process, right?”

These hearings mark the first time that I can remember any representative of the major platforms stating with so little equivocation that they can, will, and should implement steps to mitigate harmful content on their platforms.  Doubts were raised, however, by some members. Senator John Kennedy (R-LA) told Mr. Strech pointedly that he simply doesn’t believe Facebook can effectively vet over five million ads per month; and Senator Patrick Leahy (D-VT) accused all three platforms of responding too slowly, of missing opportunities and warning sings, and of hosting toxic content that is still online right now.

Although some Committee members raised concerns about First Amendment protections—in fact, Senator Ted Cruz (R-TX) cited incidents of alleged censorship of conservative views by the platforms themselves—neither any Committee Member of either party, nor any of the three witnesses, reiterated the generalization that removing illegal or harmful content from the platforms was fundamentally incompatible with the protection of free speech.  To the contrary, there seemed to be a very clear consensus that the manner in which these tools have been—and may continue to be—manipulated by bad actors is so harmful to democracy itself that, in context, free speech becomes a weapon of self-destruction. And that brings us back to that underlying theme and the real significance of what the Russian “hackers” did:  pit Americans against each other.

A New Kind of Literacy

I opined in a recent post that a new kind of media literacy is needed for the digital age.  Because no matter what Congress can legislate, and no matter what actions the platforms may take, people themselves are going to have to be more vigilant about the content they believe to be true, let alone share.  The mistaken expectation that the internet would be a kind of turbo-booster for democratic values comes from a reasonable, if somewhat elitist, assumption.  The theory was that if people have access to information, unfiltered by the influence of manipulators and monied gatekeepers, the collective wisdom of a fundamentally benevolent society would galvanize core democratic principles.  The manipulators would be powerless in such a fact-rich environment.

These assumptions completely overlooked some fundamental realities:  1) the platforms themselves can be used by a wide range of manipulators to manufacture false information; 2) false information that jibes with pre-conceived bias is almost impossible to recognize as false because; 3) people are driven more by emotion than by information.  The fault of the technologists, whose expertise is data, was to assume that information builds community—or at least to sell that idea.  But the truth is almost always just the opposite, even without propagandists hijacking reality.

The Opposite of Social Media

As an example of the limits of social media, I think about the story of Daryl Davis, who was featured by several news organizations shortly after the riots in Charlottesville.  Davis, a black blues musician, is responsible for over 200 men quitting the Ku Klux Klan—a journey that began, humorously enough, when a white man in a bar complimented him by saying he’d never heard “a black guy who could play piano like Jerry Lee Lewis.”  Davis’s explanation that Lewis learned everything he knew from black musicians led to a cordial conversation and the discovery that the white guy was a member of the KKK.  Thus began Davis’s decision to travel the country, meet other Klan members, and write a book about his experiences.  Along the way, many of the friends he made abandoned the organization and gave Davis their robes as penitential offerings.

Now, imagine if Davis’s first encounter with that first Klan member had been through the cold portals of social media rather than in person and through the shared experience of quintessentially American music. Add to that the trove of “information” the white guy could link to “proving” the genetic inferiority of Davis’s race. Can anyone doubt that the most likely outcome of this online exchange would be a hardening of the white guy’s racism (and perhaps a hardening of Davis’s feelings as well)?

Nothing in such and exchange would have to qualify as hate speech or any other content that would even get on the radar of the issues discussed in yesterday’s Senate hearing. It’s the kind of exchange that happens all the time—just two Americans being driven further apart by the mechanisms of a platform whose corporate mission is to “build community.”  That paradox is something the folks at Facebook et al—and those of us who use these platforms—are going to have to reconcile no matter what Congress does.

Image by alexlmx

Speech Maximalism on SESTA is Madness

This refrain keeps playing over in my head lately:  The EFF and its sister organizations are to cyberlaw as the NRA is to rational gun policy in America.  That seems like a pretty harsh thing to say about a bunch of progressives (and one must even include the ACLU in this discussion), but in the context of policy debate, the maximalism with which these organizations continue to defend the liability shield (Sec. 230) of the Communications Decency Act (1996) on behalf of a single multi-billion-dollar industry is logically comparable to the maximalism with which the NRA has marketed so much ahistorical nonsense about the Second Amendment on behalf of gun manufacturers.

While it’s hard to look away from the circus playing round-the-clock at the White House, it is certainly necessary to look beyond it.  The story of where American democracy is heading is not Donald Trump, though it may be (metaphorically speaking) Elon Musk.  The fact that Musk announced he could power Puerto Rico in response to official U.S. dithering is both intriguing and generous, but it is also a frightening commentary on the condition of the American state.  Even as an idea, Musk’s offer is a subtle harbinger of the tipping point I fear we may be approaching—that the state becomes so dysfunctional, the people turn to the oligarchy of technologists and say, “save us” from ourselves. At that point, American democracy will come to an end. Cue 21st-century American feudalism.

Before we head quite that far into a sci-fi thriller, though, we are truly at an inflection point when the fate of a couple of bills in Congress will say a lot about how much power and influence Google and the other major internet players have in Washington.  H.R. 1865 and S. 1693 (SESTA) would amend Section 230 of the CDA to explicitly prohibit online support of trafficking minors in the sex trade and thus open pathways to both civil and criminal prosecution.  These bills are largely a response to allegations stemming from investigations into Backpage, which the National Center for Missing and Exploited Children estimates is how 73% of all children trafficked in prostitution are bought and sold.

I am told by various contacts in D.C. that Google’s lobbyists—parent company Alphabet now ranks among the top five spenders in the country—have been out in force to kill these anti-trafficking bills in committee. Meanwhile, the EFF and other Google-funded organizations have the unenviable task of telling the American people—once again—that free speech on the web will suffer if we pass legislation designed to help protect children from sex-trafficking.  As explained in a previous post, SESTA proposes a change in the Section 230 statute that is so narrow it could never affect the vast majority of internet users.

Your site would have to be a lot like Backpage, or would have to be as big as Google or Facebook just to be in the orbit of potential liability under SESTA.  Even a pornography site that might inadvertently host video depicting sex acts with trafficked minors (and that’s a big hypothetical) would not necessarily be liable under SESTA because, depending on what actions the site owners were to take, they could still qualify for the safe harbor provisions of Section 230.  Any implication that the vast majority of us who do not run globally substantial sites, or who do not use the web to conduct transactions in the sex trade, will somehow feel a tremor in the force of free speech is rank hysteria.

But Google, with all its wealth and influence, would rather not have so much as a pinhole of liability pierced into the CDA shield—even if it means providing a modicum of legal remedy for victims of sex-trafficking by prosecuting individuals who have nothing whatsoever to do with Google. I can only imagine there must be a few members of the EFF who are either experiencing moral crises over this issue, or downing 10 a.m. shots just to quiet the cognitive dissonance because they’ve got to know their free speech arguments against SESTA are complete hogwash.

Overcoming Free Speech Maximalism

In the same way that the NRA markets a message that guns create freedom, the internet industry has sold a very similar maximalist view that the First Amendment is perpetually strengthened by the immeasurable volume of interactions on the internet.  Just as the American who owns ten guns is not ten times freer than the American who owns one gun, the American who tweets a hundred times a day is not freer than the American who doesn’t have a Twitter account at all.  Nevertheless, when one reads the declarations insisting that every peep uttered in cyberspace is sacred, it is hard to miss the rhetorical similarities between the NRA and the internet activist organizations.

Like anyone with a maximalist view—or a financial stake in espousing one—both the NRA and the EFF reveal a callous disregard for the harm being done by the policies they endorse.  The EFF hasn’t explicitly said “child sex-trafficking is the price we pay for freedom,” but that’s effectively the argument they’re making with their overplayed appeals to the First Amendment in context to SESTA.  Adding further to this irony is a complete disregard for the fact that the internet as we know it is actually making quite a hash of the democratic principles which the protection of speech is meant to serve.

In almost the same manner in which Citizens United undermines the intent of speech by giving a louder voice to financially empowered corporations, the economics of the web do the same thing more broadly and more insidiously.  If it is fundamental to American democracy that the population has access to relevant and accurate information, it is no surprise that the economics of attracting and monetizing web traffic fails to serve this purpose. (Or have I missed something and American democracy is healthier than ever?)  Journalism (i.e. information) is supposed to be the practice of telling people what they need to know while the design of the web we have is fundamentally built to tell people what they want to hear.

Adaptive algorithms that anticipate our interests, biases, and desires are relatively innocuous, perhaps even beneficial, if we’re shopping for toasters; but these designs can be toxic to democracy when we’re “shopping” for news.  In a solid, concise OpEd for Forbes about the folly of current support for Obama-era net-neutrality policies, Fred Campbell calls the internet as we know it “a mess.” “Policies that net neutrality advocates are clamoring to preserve have facilitated the internet’s roles in undermining fair elections, providing a safe haven for sex traffickers, destroying privacy, nurturing the world’s largest information monopolies (e.g., Google, Amazon), subverting free speech, and devastating publishing industries,” Campbell writes, suggesting that we should let the internet be overhauled because it’s hardly living up to the vision of its founders in the 1960s.

Campbell cites a paper by Professor Shoshana Zuboff of the Berkman Center for Internet & Society; Harvard Business School (an organization typically aligned with internet industry views), who calls the current economics of the web surveillance capitalism.  “This new form of information capitalism aims to predict and modify human behavior as a means to produce revenue and market control,” Zuboff writes.  That description certainly rings true with experience and hardly seems to jibe with the foundational assumption that the internet is “the greatest tool for democracy ever created.”

In 1783, in the uncertain period between the end of the American revolution and the establishment of the United States, Alexander Hamilton wrote to John Jay, “It is hoped when prejudice and folly have run themselves out of breath, we may embrace reason and correct our errors.”  He was referring to the many competing forces driving people away from the establishment of a unified nation.  Today, Hamilton could easily be talking about Facebook and Twitter because it would be hard to make the case that the internet is not, on balance, having a centrifugal effect on the electorate.  As such, free speech maximalism is  not only specifically immoral as a response to a bill like SESTA, but it is also generally untenable as a premise for broader debates about cyberlaw.

Image by stawy13

Fake News Tops Results After Las Vegas Shooting

On Monday, I was up early and first heard about the Las Vegas shooting on the radio in the car. It was still dark, and the winding road thick with fog, lending an eerie mood to the sound of Scott Simon’s voice on NPR reporting what little was known about this latest incident in what is now an epidemic of mass-killings. I had yet to look at any social media, to read anyone else’s opinion or to have the raw facts of the tragedy synthesized through the narrative of gun control, mental illness, terrorism, or any other matter of public policy. There was just the horrible truth of what had happened without theory or explanation. This is how we used to digest the news: Here’s what we know so far. Stay tuned.

Social media abhors a vacuum. And in the hazy interval between breaking reports of an event like the Las Vegas spree-shooting and the revelation of salient, credible details, the pranksters, trolls, and professional liars come out to play. Brianna Provenzano, writing for Mic.com, states that for several hours, “Facebook and Google’s algorithms prioritized fake news” about the Las Vegas shooting. As she puts it “conservative conspiracy sites like the Gateway Pundit lit up with misinformation about the shooter’s identity.” Her article shows one example of a headline naming some poort guy who had nothing to do with the shooting, calling him a “Democrat Who Likes Rachel Maddow, MoveOn.org, and Associated with Anti-Trump Army.”

According to Provenzano, the Gateway Pundit story was among the top results on Facebook before it was removed, but also that once the innocent man’s name was out there, Google searches for it led readers to a 4Chan thread “labeling him a dangerous leftist,” Provenzano writes. She also reports that Google eventually made algorithmic adjustments to replace the 4Chan story with relevant results and stated it will continue to be vigilant in this regard.

It’s right that Google and Facebook took action to quash, or at least mitigate, misleading “news” about such a gravely serious incident, especially bogus reports naming an innocent man as the perpetrator. But for those of us regularly following the policy positions of the internet industry, the hypocrisy here is not missed. For instance, Google can clearly take remediating steps where to no do so would look bad for them; but in other contexts in which search results may facilitate harm, they will expound ad nauseam upon the sanctity of free speech as a universal rationale to leave all data exactly where it is.

For instance, regarding the Equustek case and the Canadian court order to remove links, I fail to see a substantive distinction, in a speech context, between a counterfeiter using search to hijack customers from a legitimate product-maker and a counterfeit news-maker using search to hijack readers from legitimate reporting. In fact, ironically enough, a bogus news story, harmful and revolting as it may be in the wake of a tragedy like Las Vegas, has a better claim to speech rights than a hyperlink which leads consumers to a product or service that is breaking the law.

So, it’s not that I think Google et al shouldn’t make decisions to remove or demote “news” emanating from the adolescent babooneries of places like 4Chan. They absolutely should. Fake news is toxic, and we have enough problems with grim reality without people inventing and believing bogus narratives. But as I’ve argued more times than I can count, speech cannot be the default rationale for a universal laissez-faire policy in cyberspace. And as this story demonstrates, it’s a lie anyway. The major web platforms can and will manipulate, delete, or demote content, or links to content, when they are motivated to do so. Whether these internal decisions are driven by revenue, public relations, or even altruism, speech-maximalism does not seem to factor into their thinking, so there’s no reason why it should necessarily factor into external motivations like a court order.

Meanwhile, we can’t expect Google and Facebook to stop people from being idiots. Readers may remember that after the Boston Marathon bombing in 2013, netizens took it upon themselves to play law enforcement. Not only did they vilify an innocent man whose whereabouts were unknown, but the cyber-mob soon harassed the man’s family, who would then discover that they young man was missing because he had committed suicide.

In the early days of Web 1.0, I rejected the old cliché Don’t believe anything you read on the internet because, of course, the internet really was just a conduit, and a credible source is a credible source. But now that there’s such a bounty of absolute garbage that can either be designed to look legit or can be algorithmically elevated to undeserved prominence, that I think skepticism should be the default approach to nearly every headline. So far, the “information revolution” is at least half oxymoronic. And part of the problem is that it can be very hard to know which half.