Category: Law & Policy

Section 230: Fix It or Lose It?

In 2015, Rep. Jackie Speier (D-CA) introduced a bill that would make it a federal crime to engage in what is generically called “revenge porn.”  I say generically because “revenge” alludes to a specific motive, usually that of a disgruntled ex-boyfriend who decides to get back at a former girlfriend by distributing intimate or sexually explicit images of her online.  There are revenge porn websites dedicated to hosting this type of material, and in some cases, site operators have engaged in extortion, demanding money from victims in exchange for removing their images. 

Naturally, the usual suspects responded to Rep. Speier’s proposal with the usual hand-wringing jitters, asserting that any implication of platform responsibility for almost anything will only lead to eroding the proper functioning of the internet.  (Is it functioning properly?)  As quoted in my 2015 post, Mike Masnick at Techdirt stated, regarding the Speier bill, “Trying to accurately describe what ‘revenge porn’ is for the sake of criminalizing its posting, will almost certainly have chilling effects on third parties and undermine the very intent of the CDA’s Section 230.”  [Emphasis added]

But legislation like this does not undermine the intent of Section 230 of the Communications Decency Act, and saying otherwise grossly misrepresents—in fact inverts—the goal of that liability shield when it was written into law in 1996.  Section 230 of the CDA was specifically enacted to encourage content moderation by platform owners to remove unlawful or harmful material.  Unfortunately, this “safe harbor” provision has since been reimagined by the internet industry, web activists, and some jurists as a legal foundation to avoid content moderation—even when ordered to do so by a court of law.  That is an aberration of what CDA230 was meant to achieve.

To date, forty-one states plus Washington D.C. have criminalized non-consensual use of sexually explicit or intimate visual material, and New York is poised to join this company with a new bill now proceeding through the State Assembly.  Notably, the language in this bill (similar to Speier’s federal proposal) suggests to me that identifying the criminality of this particular conduct is not so far outside the scope of legislative capacity as Masnick implied in 2015.  The New York bill states, “…with intent to cause material harm to the emotional, financial or physical welfare of another person …”  That doesn’t seem very complicated.  If the goal is to hurt someone, regardless of why, then criminal conduct may be present.  

Of course, the tech pundits don’t really mind criminalizing the behavior of the individuals who commit “revenge porn.”  I won’t accuse Masnick, the EFF, et al of supporting the people who engage in this type of conduct because they certainly do not.  What they do claim to be concerned about are the broader implications for internet platforms if they can be held liable under the criminal code, or even just directed by court order to remove material as a form of injunctive relief for victims.  Here, the critics rely on the well-worn generality that any gap in the great wall of Section 230 will only result in reactionary responses by well-meaning web platforms, which will then censor otherwise protected speech.  

Maybe I lack imagination, but it is actually impossible to fathom how providing a relatively narrow path to legal remedies for the victims of this singular crime can chill anything related to the normal functioning of most online activity. Someone wins a revenge porn case, and what?  We won’t be able to read the New York Times or buy sneakers on Amazon or watch Hulu?  Bullshit.  

As usual, the pundits tend to overlook the fact that due process is still required—that an alleged victim still has to prove her case and demonstrate how a named platform may be criminally or civilly liable for harm.  And in many cases, a platform may be responsible for nothing more than removing content without facing any further liability whatsoever.  Meanwhile, people have already been held criminally and civilly liable for various types of revenge porn, and material has been removed from various site, and the internet is still functioning.  In fact, one audacious law firm in Brooklyn, NY focuses on exactly these issues under the direction of attorney Carrie A. Goldberg, who says she became the lawyer she needed herself after an ex-boyfriend threatened to post naked pictures of her online.  

Since then, Goldberg’s firm has removed over 20,000 images on behalf of its clients—a number that simultaneously demonstrates there is efficacy in criminalizing non-consensual uses of material, but it also proves my point about due process and the narrowness of this focus.  In short, the socially-beneficial aspects of the internet really can endure the removal of many thousands of illegal or harmful files without the rest of us feeling a thing, and it is preposterous to believe otherwise.  Or as part of Goldberg’s Twitter bio explains:  F*uck your overbroad reading of CDA230.

I liked that slogan so much, I asked her for coffee mug with the words printed on it.  I guess it’s just the kind of nerd-moxie that makes my day, and Golberg’s firm certainly appears to have moxie to spare, as exemplified by this declaration on their website …

We are done living in a world of abuse and we are not afraid to sue the *&%$ out of schools, tech companies, and employers who tolerate it.  There are many ways to get justice for our clients – economic justice, restraining orders, advocacy in Campus Disciplinary proceedings, exposing a predator, getting the piece-of-shit thrown in jail.

To the extent this take-no-prisoners attitude accurately sums up a general shift in public sentiment (i.e. that some form of platform responsibility is mandated), I suspect the whinging chorus of internet activists may soon need to find a new cross to die on other than their adamantine devotion to the sanctity of Section 230.  In fact, it is conceivable that if the tech giants do not get on board and help tweak—or at least don’t stand in the way of tweaking—the application of this liability shield, they just might lose it altogether.

Apropos my last post about the implications of deepfakes, this universe of criminal conduct will likely become more complicated as parties willing to cause harm can more easily manufacture visual material that appears to reveal the intended target(s) engaged in embarrassing, or even illegal, activity.  For instance, most, if not all, of the revenge porn statutes criminalize visual material that depicts the “intimate parts” of the plaintiff bringing a claim, and this language would seem to fall short of criminalizing a deepfake in which the victim’s face has been seamlessly grafted onto someone else’s body.  Hence, the criminal codes may already be lagging behind the technology.

And, of course, the implications here are much broader than non-consensual pornography.  Just look at the consequences (in this case almost certainly deserved) for Virginia Governor Ralph Northam over a 1984 yearbook in which he appears at least adjacent to, if not depicted in, racist and demeaning photographs. Those photos are real, and Northam must deal with the consequences, but we are now well past the point when far more sophisticated imagery than yearbook photos can be fabricated out of thin air by someone with rudimentary skills.  Combine the level of destruction that can be so easily achieved with precedent application of Section 230 (e.g. Yelp refusing to remove a handful of libelous reviews), and it seems to me that change is coming, and the big platforms may want to get on board.  

As I posted last July, a new development in this narrative—and one I consider unfortunate—is the addition of partisan politics to the mix.  Some conservative Republicans in Congress have at least hinted at eradicating Section 230 in response to allegations that web platforms promote left-leaning content over right-leaning content.  Clearly, this specific complaint implies a tangled mess of a debate that nobody should want; but if the legislative Venn diagram encompasses those who want to kill 230 with those who want to carve out reasonable remedies for online harm, Google and Co. may need to change their revisionist narrative on the purpose of that liability shield, or risk losing more than symbolic battles.    

Techdirt Dissembles on FOSTA in Rebuke of Kristof

In a recent post on Techdirt, Mike Masnick calls columnist Nicholas Kristof a hypocrite based on a narrative Masnick just plain made up.  On December 12, Kristof published a brief column in The New York Times with a picture of a 12-year-old girl who is starving to death as a victim of the US-backed, Saudi-Arabian war in Yemen.  The girl is naked but for a diaper and a bandage on her foot, and the image of her skeletal, wasting body is truly humbling, which is why Kristof says he devoted so much of the page to the image itself.  

After the story was posted on Facebook, the social platform apparently kept deleting the photograph, which prompted the following tweet from Kristof on December 16:

Facebook seems to have repeatedly blocked the photo of Abrar that went with my column:  Come on, Facebook! If you want to end these horrifying images of starving children in Yemen, then help end the U.S.-backed Saudi war that causes the starvation.

Kristof’s complaint was then seized upon by Masnick, who concocted a typically sarcastic “gotcha” on the premise that because Kristof backed the counter-sex-trafficking legislation known as FOSTA (Fight Online Sex Trafficking Act), he has no right to “whine” about Facebook removing this photo for its “sexual content.”  True to form, Masnick smugly alleges that Kristof knew nothing about how FOSTA worked, despite the fact that Masnick grotesquely misrepresents the law, as well as the nature of Facebook moderation, in his post.

Without even getting into FOSTA, anyone who has been on Facebook for the past decade or so knows that the platform has often removed images—even fine art—that some moderator believed violated its “community standards.”  Facebook has been making these, often laughable, mistakes since long before anyone introduced the legislation that became FOSTA and which passed into law in April of this year.  In fact, Masnick’s recent post cites one of his other posts from 2016 criticizing Facebook for censoring the iconic, Pulitzer Prize winning photo of the naked Vietnamese girl running from a napalm strike.  

Notably, the removal of that famous photograph was actually mentioned in the documentary The Cleaners, which I wrote about in November, and which profiles the Philippines-based  moderators to whom Facebook has outsourced most, if not all, of its “community standards” oversight.  The documentary reveals a melange of human fallibility in the decision-making behind content moderation, and Kristof’s photo might have been repeatedly removed for being “disturbing” rather than “sexual.” 

Regardless, the broader point is that millions of images a day are processed by these young moderators—and they are required to meet quotas—whose culture is not grounded in American principles of speech, press, etc., and it is almost impossible to generalize about their motivations and judgment calls.

At the same time, even if, in the most depraved imagination, someone could identify Kristof’s photo of this poor child as “sexual,” then it would simply violate child pornography laws, which predate FOSTA, predate Facebook, and even predate the birth of Mark Zuckerberg.  Yet, somehow The New York Times published the image, which nobody seems to have confused with pornographic exploitation.  All of which is to say that neither the Facebook moderation regime nor Kristof’s specific complaint about the photo, which no sane person could confuse as “sexual,” has anything to do with FOSTA.

As explained in several posts, what FOSTA does is affirm that no internet service provider is automatically immunized against criminal or civil allegations of contributing to sex-trafficking.  FOSTA does not mean that a plaintiff who brings a claim has any less burden to prove a platform’s culpability in that crime.  (Y’know, the way the law works.)  In fact, all one needs to do is look at the volume and nature of the evidence gathered against Backpage to see that proving a contributory role in sex-trafficking takes a hell of a lot more than hosting some “nudity.”

While it is possible that, in an abundance of caution after passage of FOSTA, attorneys at Facebook recommend simply removing anything that can even remotely be deemed “sexual,” it is also evident that the platform was generally doing this long before FOSTA.  Next, the platform will, and should, remove material that is patently child pornography.  And finally, the attorneys at Facebook are well aware that hosting content which may be used as evidence of “contributing to sex-trafficking” is a distinct and high bar for a would-be plaintiff to meet.

So, it is a leap and a half to allege that platforms are now over-censoring as a result of FOSTA, to say nothing of the current reality that Facebook has way bigger content moderation problems right now.  In this regard, I think the folks at Techdirt, and everyone else, ought to be more concerned that Facebook cannot seem to distinguish between a third-party like The New York Times and just some other account holder.

It ought to be a simple enough, internal practice to determine that if a mainstream news company—which is also not immunized against allegations of illegal conduct—can publish an image without legal jeopardy, then Facebook can safely host the same image.  Why this does not appear to be the case has everything to do with the platform’s overall management and nothing to do with FOSTA.  

I’ll leave it to the judgment of the reader to consider Masnick’s labeling Kristof as having a “savior complex” for his interest in starving children and trafficking victims.  But given the choice between a guy who wants to save kids and a guy who wants to save legal liability shields for mega-corporations, well, let’s just say Mike may not make the Nice list this Christmas.

Internet Platforms Above the Law?

Silicon Valley may have done ‘bare minimum’ to help Russia investigation, Senate Intel Committee told … 

That headline from CNN, and which was echoed in several news stories that began appearing late Monday, will elicit no surprise among my friends and colleagues working in IP law, privacy, publicity rights, security, and various other matters of justice in the digital marketplace.  Those of us who believe that the rule of law is not anathema to the internet are used to the major platforms behaving as though they operate in some alternate universe where the laws of old-model, physical humans do not apply.  After all, these companies and their executives were nurtured on exactly that manifesto when the late John Perry Barlow first read his Declaration of the Independence of Cyberspace at Davos in 1996. 

To put it mildly, Barlow was utterly full of shit when he declared the internet to be a “home of mind” that could not, or should not, be governed by the “weary giants of flesh and steel.”  Perhaps he can be forgiven some poetic license in the service of a sincere hope that the internet might truly be an incorporeal space that would help us transcend human folly and connect to one another through our better angels.  But that’s not what happened.  And debates about cyber policy should stop pretending it can still happen.

Fast-forward to the harsh realities of the present, and the “home of mind” is dominated by glorified advertising platforms, easily exploited by the worst kind of malicious actors and which clearly appeal to our lesser—even our profoundly stupid—angels.  And the sad irony is that, far from operating benignly adjacent to physical society, social media platforms have been exploited to infiltrate, corrupt, demoralize, and degrade the foundations of society in very real and very dangerous ways.

Two independent reports commissioned by the Senate Select Committee on Intelligence investigated the St. Petersburg-based Internet Research Agency (IRA), the professional troll farm erected for the sole purpose of inflaming political discord in the United States and other liberal democracies around the world.   The reports reveal that disinformation on all major platforms was, and still is, more widespread than initially believed; and they describe the methods by which specific groups like African Americans were targeted in an effort to dissuade voter turnout for the 2016 election.

I plan to read both reports and follow up, but for now, I thought it worth highlighting the detail that the reports’ authors allege that the major platforms were far less cooperative than one might hope given the gravity of the circumstances.   As The Washington Post quotes …

“Unfortunately, Google made the unusual choice to provide data to the Committee in nonmachine‐readable format.  The ads data was provided in lengthy PDF format whose pages displayed copies of information previously organized in spreadsheets (Google could have provided the original spreadsheets in CSV or JSON files).”

Think about that one.  Google stifling the use of computers as the U.S. Senate tries to better understand exactly how a foreign and hostile power has been working to fracture the American democratic process.  Further, while skimming the report submitted by New Knowledge, I caught the statement that begins, “Regrettably, it appears that the platforms may have misrepresented or evaded in some of their statements to Congress.”  

In this regard, I was intrigued by the strident tone lately adopted by Senator Ron Wyden in response to Silicon Valley’s less than forthright conduct in these investigations.  Vowing to pass “legislation with teeth,” Wyden has proposed a new consumer privacy bill aimed at restricting what these platforms may do with user data, particularly with respect to the manner in which that data may be leveraged to target disinformation about politics and policy issues.  Further, the proverbial “teeth” in Wyden’s bill would impose substantial penalties, including potential imprisonment for executives, for failing to provide honest and complete testimony to Congress.

For my colleagues who work in copyright advocacy, Wyden has not exactly been the rule-of-law representative when it comes to holding Silicon Valley accountable.  But perhaps the thinking will change as the senator and his colleagues must now address the many indisputable ways in which a liability-free internet industry has, quite possibly, done more harm than good for American democracy.

Meanwhile, despite mounting evidence that the major social platforms are more often a home of mindlessness than mind, Barlow’s Declaration remains the cosmic background noise still ringing in the heads of too many defenders of what we generically call “the internet.”  Whether it’s the Electronic Frontier Foundation or Techdirt or MEP Julia Reda’s anti-copyright campaign in the EU or the Internet Association or even the American Library Association, one can still hear the strains of a misguided faith in a pure internet, unsullied by the taint of law, in the rhetoric deployed against almost any policy that might demand platform responsibility.

For far too long, a false premise undermining copyright enforcement specifically—and almost all other types of enforcement generally—has been that it is better to allow harmful or illegal content to remain online than to risk censoring even a micro-byte of protected speech.  But that premise is, paradoxically enough, a pre-digital-age idea and a blind allegiance to Barlow’s naive cyber-utopianism.  It is a laissez-faire approach, which casually ignores the new reality in which an unfettered amount of harmful or illegal content continues to undermine the values it claims to uphold.  

After all, if an American inadvertently shares a political meme that was written by a malicious actor in St. Petersburg—and whose goal is to weaken global democracy—can anyone honestly say that free speech is fulfilling its purpose in that moment?  This is just one reason why, about a month ago, I personally stopped most activity on Facebook:  because I’m not sure it’s possible to avoid feeding that particular cancer.  At the same time, it is notable that we do not even need bad actors as dramatically well-organized as the IRA to weaponize disinformation on a wide range of issues; but that’s a subject for a different post. 

So, I look forward to reading the Senate-commissioned reports; but for now, I thought it worth noting the allegations that the major platforms are stonewalling and obfuscating in these investigations—still behaving as though they operate outside the rule of law.   Of course. the tragically ironic twist to Barlow’s vision of cyberspace as separate from, and elevated above, “weary” reality is that our present reality too often resembles the craven, mean-spirited, and willfully misinformed cyber-world of social media.

Robot image source by digitalstormcinema