TABLE OF CONTENTS I. INTRODUCTION 279 II. THE STATE OF FAKE NEWS: THE FIRST AMENDMENT AND [SECTION] 230 283 A. The Protections of the First Amendment 283 B. The Shield of [section] 230 287 III. REGULATION BY EDGE PROVIDERS 291 IV. LESSONS FROM ONLINE COPYRIGHT LAW 294 A. The Digital Millennium Copyright Act 295 B. Courts on Vicarious Liability 298 C. Shortcomings of the DMCA 300 D. Lessons for Fake News 301 V. CONCLUSION 305 I. INTRODUCTION
Fake news is hardly new; it has long been a common tactic of politics to shift the truth and ignore questions. (1) Indeed, the publication of fake news stretches back to the birth of the printing press, if not earlier, (2) but the rise of the Internet and social media has fundamentally changed the possibilities around truth and fake news. The vast majority of Americans get at least some of their news from the Internet. (3) Seven in ten Americans use social media. (4) At the same time, traditional journalism has declined, and there are fewer reporters despite a proliferation of thousands of online news sources and blogs that still need content to fill them. (5) Now the marketplace is full of false information that is packaged to look true. (6) Compared to traditional speech via word of mouth, print, or broadcast, the dissemination of information over the Internet is distinct in that it combines content filters, insular online communities, amplification of fringe ideas, the rapidity of idea dissemination, and profit incentives that encourage fake news. (7)
Even if fake news has existed for centuries, the modern concept is far different and more dangerous. In the past few years, we have become inundated with purposefully adopted false information that quickly and efficiently seeps through society. (8) For example, misinformation likely played a significant role in the outcome of the 2016 presidential election. (9) In the lead-up to the 2016 election, a BuzzFeed News study found that fake news stories generated more user activity than did legitimate news stories by well-reputed sources like the New York Times and the Washington Post. (10) Pope Francis endorsing President Donald J. Trump, (11) Hillary Clinton running a sex trafficking ring from a pizza parlor, (12) and Clinton selling weapons to the Islamic State (13--)all are false, but were widely shared online. In the aftermath of the 2016 election, fake news has continued to grow. (14)
Fake news is not limited to the electoral realm either: "The harmful information that spreads on Facebook includes the myths and lies about vaccination and links to autism. It contains myths and lies about the scientific fact of global warming. These are issues that are crucial to our wellbeing." (15) One recent incident of fake news even claimed that the Turkish army had invaded and occupied Greek territory across the Evros river, which separates the two countries. (16) The World Economic Forum went so far as to conclude that online misinformation creates "digital wildfires" that pose a global problem. (17)
By removing the necessity of a publisher, the Internet increased opportunities to directly share content with a vast audience. (18) Two of the primary reasons for such articles are the purposeful spreading of fake news to support a position or sow chaos and the creation of profit-generating clickbait. (19) Some of these peddlers of fake news stories were "fake news websites that only publish hoaxes or... hyperpartisan websites." (20) In addition, fake news pays--enticing headlines, even if false, generate traffic, which in turn drives increased advertising revenue as more people visit the website. (21) One such story--whose headline read "BREAKING: 'Tens of Thousands' of Fraudulent Clinton Votes Found in Ohio Warehouse"--was shared by six million people and generated $5000 in advertising revenue for the creator and poster of the story through Google advertisements on the story's website. (22) Additionally, it is significantly more expensive to create true stories, which require reporting and research, than false ones. (23)
Fake news is undoubtedly a problem in the United States, yet there are few legal constraints to stop it. Individuals who post fake news are immunized under the First Amendment. (24) The Communications Decency Act ("CDA"), codified as 17 U.S.C. [section] 230, prevents websites that host users' fake news from being held accountable. (25) The First Amendment and [section] 230 thus operate in tandem to provide a significant shield for fake news, protecting it as free speech and protecting hosting platforms from liability for defamation. In response to this framework, scholarship on fake news has focused on how the First Amendment and [section] 230 create a powerful shield for fake news, (26) how the First Amendment and [section] 230 regime could be modified, (27) how online content providers should regulate fake news posted on their websites rather than the government, (28) and how online content providers are self-regulating. (29) This Article, instead, will focus on a pre-existing model for regulating fake news on websites, the Digital Millennium Copyright Act ("DMCA"), and will discuss principles from the DMCA that could be expanded into the fake news context, whether for website self-regulation or future federal law.
Scholarship that contemplates using the DMCA to regulate fake news has largely focused on government regulation. For example, Lee Royster argued that the DMCA notice and takedown regime should be used by courts in imposing liability for fake news. (30) Alternatively, Emma Savino argued that the [section] 230 regime should be modified by Congress to hold edge providers (31) responsible and suggested the DMCA notice and takedown regime as a possible model. (32) Similarly, Andrew Schuyler suggested that the DMCA could provide a useful model for imposing liability on edge providers for libel posted by third parties, even suggesting a congressional bill modeled after the DCMA for this purpose. (33) This Article will instead examine the utility of DMCA principles for both government regulation and self-regulation of fake news by edge providers, and will pull not only from the DMCA itself, but also case law and the recently released Copyright Office report on the DMCA's efficacy.
In Part II, this Article will lay out the current legal framework for fake news on the Internet under the First Amendment and [section] 230. Part III will discuss how websites are starting to self-regulate, looking at the particular mechanisms they are adopting. After laying this groundwork for fake news regulation, this Article will then turn to the DMCA as a model for regulation of proscribed online activity that holds websites accountable. In Part IV, this Article will first describe the confines of the DMCA, related case law, and the Copyright Office's recent findings on [section] 512 efficacy. It will then discuss ten principles from this law that could serve as guidelines for regulating fake news, whether by websites themselves or through government action. Finally, Part V concludes and looks to the future.
THE STATE OF FAKE NEWS: THE FIRST AMENDMENT AND [section] 230
The Protections of the First Amendment
First Amendment protections for free speech present an overwhelming problem for the regulation of fake news. (34) Speech on the Internet enjoys "the same level of constitutional protection as traditional forms of speech." (35) The Supreme Court has said that false statements are less valuable than true statements and that this gives them less protection under the First Amendment, (36) but it has only said this in the context of defamation or other legally cognizable harms. (37) For pure content where there is no cognizable or provable harm to an individual, the First Amendment's goal is the protection of speech--whether true or false--and this principle has been upheld repeatedly by the Supreme Court. (38)
Overall, free speech is broadly protected under the heightened strict or intermediate scrutiny standards elucidated by the Court. (39) To meet strict scrutiny, the government must show that its restriction on speech is narrowly tailored to meet a compelling government interest. (40) To meet intermediate scrutiny, the government must show that its restriction is substantially related to an important government interest. (41) But where speech is unprotected, Congress has much more leeway. (42) The Supreme Court has generally defined the confines of the First Amendment by carving out certain types of unprotected speech, such as obscenity, (43) fighting words, (44) child pornography, (45) and incitement of imminent lawless action. (46) However, such carve-outs have been limited, and the Supreme Court has generally been guided in First Amendment cases by a fear of allowing the government to become the ultimate arbiter of free speech. (47) Even if the rigidity of heightened scrutiny has been criticized as weakening over time, (48) these are still high bars to meet. This is the rationale for why the Supreme Court has repeatedly upheld the protection of false speech under the First Amendment. (49)
Most recently, this principle was upheld in 2012 in United States v. Alvarez, (50) where the Supreme Court held that it was unconstitutional to criminalize making false claims to military honors. (51) The Court noted that its chief concern was not that one cannot tell the difference between truth and lie, but that the ends of free speech and discourse "are not well served when the government seeks to orchestrate public discussion through content-based mandates.... Only a weak society needs government protection or intervention before it pursues its resolve to preserve the truth." (52) Although this was a plurality opinion, all of the Justices appeared to be skeptical of giving the government the authority to broadly police false speech. (53)
The First Amendment protections of false information have been largely premised on...