Anthony M. Kennedy Professor of Law & Executive Director, Capital Center for Law & Policy, University of the Pacific, McGeorge School of Law, Sacramento, California, USA.
The American Journal of Comparative Law, Volume 70, Issue Supplement_1, October 2022, Pages i278–i311, https://doi.org/10.1093/ajcl/avac010
02 June 2022Leslie Gielow Jacobs, Freedom of Speech and Regulation of Fake News, The American Journal of Comparative Law, Volume 70, Issue Supplement_1, October 2022, Pages i278–i311, https://doi.org/10.1093/ajcl/avac010
Navbar Search Filter Mobile Enter search term Search Navbar Search Filter Enter search term SearchThis Report, prepared for the 2022 General Congress of the International Academy of Comparative Law, examines the United States law regarding freedom of speech and regulation of fake news. Freedom of speech refers to the limits the Supreme Court’s interpretation of the individual right sets on the types and scope of regulation by government at all levels of fake news. Within these limitations, regulation of fake news is critical, in the United States and around the world, as its spread undermines the functioning of the democratic processes that underpin the Constitution and subverts the value of an unrestricted and robust marketplace of speech that guides the Court’s interpretation of the free speech guarantee. In the age of the internet, fake news shoots at lightning pace around the globe, influencing opinions and actions in a matter of minutes.
The interconnectivity of communications makes regulation of fake news by nations a particularly worthwhile area of comparative law study. Nations may attempt to regulate fake news but find themselves unable to stop the flow of unwanted speech generated from locations without such restrictions across their borders. Speakers, from individuals to massive internet service providers and social media platforms, may find that when the speech they generate or host crosses borders, different rules apply, and they may be liable for harms caused by their fake news speech, even though the laws of their home venues protect them. Comparative study makes the different scopes and details of laws regulating the global problem of fake news transparent and aids collaboration among regulators from different nations, expert external entities, media speech hosts, and other stakeholders to develop and implement laws and best practices to address it.
The purpose of this Report is to set out the constitutional framework that structures regulation of the broadcast and spread of fake news in the United States, to describe the types of fake news laws that currently exist within this framework, and to note new laws and proposals aimed specifically to address fake news and its adverse public consequences. The United States is large; the system of federalism set out in the body of the Constitution diffuses authority to implement laws to protect against the harms caused by the spread of fake news through the federal and state governments, and issues surrounding the problem of fake news and responses are evolving daily. Therefore, this Report will examine the types of approaches used by the federal and state governments to address the problem of fake news in various settings and give illustrative examples. It will not be comprehensive in identifying all areas of regulation that may involve fake news or in listing all laws within the areas identified. Readers interested in more in-depth coverage should consult authorities within the topics addressed.
“Fake news” as a phenomenon spans the globe and dates back centuries. 1 As the source of focused concern in the United States, however, fake news as a concept is quite modern and, even so, is ill-defined. Watchdog websites used the fake news label before the 2016 presidential election to identify misinformation, meaning false information, in news stories and other places across the internet. 2 Their research uncovered the fake news subset of disinformation, meaning falsehoods deliberately disseminated to serve the purposes of the distributors. Then, the 2016 election moved Donald Trump into the White House, who popularized the fake news label in a third variation as rhetoric describing information offered by journalists or news organizations he viewed as enemies for the purpose of undermining their credibility. This third variation coopts the descriptive term as advocacy against the possibility of non-biased journalism, or, in a more extreme version, against the possibility of objective truth and becomes an item of misinformation or disinformation itself. 3 Misinformation, disinformation, and the dishonest labeling of truthful news stories as fake news are harmful and, in many applications, overlap and interact symbiotically. “Fake news” as the focus of this Report thus encompasses all three variations in meaning. This Report will hereinafter refer to misinformation as the most inclusive term.
Recent American experiences with the spread of misinformation identify the components of the phenomenon, which signal the appropriate targets of regulatory interventions. First, there is the speaker. Russian operatives intervened in the 2016 presidential election by planting false information on social media. Second, there is the media platform. Operatives relied on platform policies and technologies, such as paid-for “boosted advertising,” human-simulating bots, and amplifying algorithms, to spread the false speech. Third, there are the recipients—so-called “useful idiots”—who view the claims and spread them, wittingly as false or unwittingly. While the label is hardly flattering, it appropriately identifies speech recipients as crucial participants in the spread of misinformation and appropriate targets of regulatory interventions to stem it. A small investment by the Russian instigators yielded hundreds of millions of shares. 4
The Russian 2016 election intervention provides but one illustration of the speakers/platforms/recipients’ interaction that characterizes the modern misinformation phenomenon. Speakers who initiate misinformation come in all types, including government officials or candidates for office, corporations, advocacy groups, and lone individuals. Although social media has been the communication platform of focused concern, traditional media platforms such as print and broadcast outlets, and content providers such as television networks, also host misinformation and amplify it. And, of course, recipients of speech that travels through these forms of media come in all the many varieties of persons who exist across the nation, different in so many ways, but the same, with respect to misinformation, in their experience of its harms and complicity in its spread. Following a brief discussion of the limits the U.S. Constitution places on regulation of false speech, these components that facilitate the spread of misinformation—speaker, platform, and recipients—understood to include the broad range of applications, organize the sections that set out existing and proposed regulations to combat the phenomenon.
Regulation to stem the spread of misinformation must occur within the complex structure of the U.S. Constitution’s free speech guarantee. Regulations may take the form of restrictions of false speech. But, because the Constitution severely limits the government’s ability to aim at the content of speech so directly, the government must often turn to less direct means, such as requiring speakers to add information to their speech or regulating media platforms to encourage them to broadcast diverse viewpoints or not to broadcast misinformation. Because these indirect efforts involve regulating truthful speech, the entire, and complex, free speech structure provides important background to the government’s authority to regulate to prevent the harms caused by false speech. This Part will provide a brief overview.
The Free Speech Clause provides, “Congress shall make no law. . . abridging the freedom of speech, or of the press.” 5 Although the words say “Congress,” the United States Supreme Court has held the free speech guarantee to restrict the actions of all government agencies and officials, across the branches of government—legislative, executive, and judicial—and through the levels of government—federal, state, and local. 6 “Speech” includes information spoken and written, art, music, symbolic expression, and any other form of communication across all types of media.
`The free speech guarantee implements the values that underpin it, which include facilitating representative democracy and self-government, advancing knowledge and truth in the “marketplace of ideas,” and promoting individual autonomy, self-expression, and self-fulfillment. 7 These values appear in constitutional doctrine as a sharp distinction between government regulations of speech because of its content and speech regulations that are content-neutral, restricting the time, place, or manner of speech without respect to the speaker’s message. 8 Content-based speech restrictions presumptively violate the free speech guarantee because they suggest that the government is censoring information or opinions that it does not like and “raises the specter that [it] may effectively drive certain ideas or viewpoints from the marketplace.” 9 Nevertheless, some speech restrictions may overcome this presumption. The Court has adopted a categorical approach to defining the circumstances in which the government may impose content-based restrictions on speech. 10 Each category corresponds to a particular harm that the government seeks to avoid by restricting speech. The definition of each category identifies the features of speech the government must prove to exist to demonstrate that the speech restriction is consistent with the free speech guarantee. Additionally, and quite notably, the presumption that content-based regulations violate the free speech guarantee applies to the types of speech that the Court has interpreted to fall within it. 11 The reality is that “much, perhaps most, human behavior take place through speech” 12 and the Court has presumed most speech that occurs in everyday conversations and much speech that occurs within specific realms of government regulation to be outside the scope of speech protected by the Constitution’s strong presumption against content-based restrictions. 13
Outside the defined categories and inside the scope of full protection by the free speech guarantee, content-based restrictions of individual speech are valid only if they pass strict scrutiny, meaning the restriction is narrowly tailored to fulfil a compelling purpose. 14 Content-neutral speech regulations pose a lesser threat to core constitutional values and are valid so long as the means, e.g. noise-level or time-of-day restrictions, directly and materially advance the government’s substantial purpose. 15 Narrow tailoring requires that a law not be overinclusive or underinclusive, and that no alternate means other than suppressing the speech exist to fulfill the government’s objective. Adding more speech to public discourse is an alternate means to avoid the harm of speech, expounded as the preferred remedy in an early influential opinion and frequently relied upon by the Court as a ground to invalidate speech restrictions. 16 In combination, these elements of narrow tailoring are extraordinarily difficult to meet. 17
Doctrine expanding the scope of the free speech guarantee poses additional limitations to government regulation of speech. During the last several decades, the Court has been expanding the scope of speech restrictions that are content-based and trigger strict or heightened scrutiny, 18 and the Court has made clear that it is not inclined to create new categories of unprotected speech. 19 It has condemned viewpoint discrimination as a particularly egregious form of content-based regulation, 20 and has been alert to identify it as the government’s motivation, even when the law does not explicitly classify speakers this way. 21 The Court has held that the payment of money for speech receives the same constitutional protection as speech itself, and invalidated content-based campaign finance restrictions, reasoning that the Constitution prohibits the government from regulating the expenditure of money for the purpose of equalizing, or otherwise modifying the relative abilities of speakers to participate in election speech. 22 It has held that the Constitution protects corporations’ non-commercial speech to the same extent that it protects the speech of individuals, and struck down regulations imposing special restrictions on corporations’ and unions’ expenditures on election-related speech. 23 The Court has held that requirements that speakers broadcast public issue messages prepared by the government or offered by other private speakers provoke the same presumption of unconstitutionality as speech restrictions, although it has upheld certain sponsorship disclaimers on election advertising. 24 It has interpreted the Constitution to protect the previously unprotected category of commercial speech and has been tightening the scrutiny it applies to restrictions, emphasizing that it will not accept “paternalistic” justifications by the government for restricting broadcast of truthful advertising, 25 and contracting the authority of governments to require product vendors and service providers to provide more information to consumers along with their own advertising. 26
With respect to false statements, case law distinguishes between statements of ideas, which cannot be proven false and therefore cannot be restricted, and statements of fact, which can be verified. 27 A statement in the form of an opinion may be restricted by law if it implies a false statement of fact. 28 Parody and satire are protected forms of speech. 29 The Constitution does not exclude false statements of fact as a general category from protection under the free speech guarantee. 30 In making this clear, the Court emphasized that even intentionally false statements of verifiable fact may have value in personal conversation or public discourse, and that permitting the government to punish speakers simply because they utter a false statement would present dangers of government overreach that the free speech guarantee guards against. 31 These dangers include anointing the government as a type of “truth police,” empowering the government to engage in selective prosecution of speakers with viewpoints critical of it or that it otherwise dislikes and chilling the expression of protected speech. 32
To regulate false statements because they are false, the government must link its restriction to preventing a harm that it has the authority to avoid. Even with such linkage, whether the Constitution protects false statements of fact from regulation depends upon the circumstances and the terms of the regulation. 33 The Court articulated this rule in split opinions and so the nature of the harm and the tightness of the linkage are not clearly defined. 34 Judicial review of future applications will add precision. Existing applications and statements by the Court about considerations and values relevant to the constitutionality of regulations of false speech provide guidance.
Some instances of permissible restrictions are apparent. Presumably, the government may restrict false statements that occur within the realms of regulation in which the Court has tolerated content-based restrictions greater than could constitutionally be imposed on fully protected speech. The Court has said repeatedly, although not recently, that false or misleading commercial speech falls outside the free speech guarantee, so the government may restrict false statements of fact or require additional information in commercial advertising and product labels to prevent consumer confusion or deception. 35 Fraud is a category of unprotected speech. 36 Fraud may occur in the form of false commercial advertising but is more broadly defined as false representations of matters of fact that deceive and are intended to deceive others so that they will act upon it to their legal injury. 37 Most recently, the Court has affirmed that many criminal false statement laws that require proof of intent and are narrowed to avoid harm or significant risk of harm to government entities or processes are consistent with the free speech guarantee. 38
On the flip side, some limits on restriction of false speech exist. In New York Times v. Sullivan, the Court interpreted the Constitution to impose limits on common law rules of defamation damages liability, reasoning that that “erroneous statement is inevitable in free debate,” and pulling within its protection certain merely negligent false statements of fact to provide “breathing space” for the expression of truthful speech about public issues. A person who is a government official or public figure may recover damages for injury to reputation only by showing that a statement was false and proving by “clear and convincing” evidence that the speaker acted “with ‘actual malice,’ meaning knowledge that it was false or with reckless disregard of whether it was false or not. 39 A version of the heightened proof requirements apply, as well, to private people who become “limited public figures,” involved in particular public issues. 40
The “actual malice” limit on liability applies to other torts that do or may impose damages liability for false speech that occurs as part of public discourse. One of these is the “false-light” invasion of privacy tort, where the harm consists of inflicting emotional distress on a person by publicly portraying them in a false, and offensive, way. 41 Another is intentional infliction of emotional distress, which may, but does not necessarily, hinge on false statements. 42 In these cases, as in the defamation precursors, the Court has emphasized that the “threat to the free and robust debate of public issues,” the “potential interference with a meaningful dialogue of ideas,” and the “‘reaction of self-censorship’” on matters of public import that damages liability would provoke explained why the Constitution protects the harmful speech. 43
The Court has also held that speech by candidates during elections requires the “breathing space” afforded by the “actual malice” liability threshold. Reviewing application of a state statute that would have removed a winning candidate from office for a false statement made in good faith and quickly retracted, the Court noted the “chilling effect [that] such absolute accountability for factual misstatements in the course of political debate” would impose and observed that because “[i]n a political campaign, a candidate’s factual blunder is unlikely to escape the notice of, and correction by, the erring candidate’s political opponent. . . the preferred First Amendment remedy of ‘more speech, not enforced silence,’ thus has special force.” 44
When false statements of fact involve issue of public concern, even requiring that the speech be intentionally false may not be enough to satisfy the free speech guarantee. Strict scrutiny requires that laws limiting speech be drafted narrowly so as not to deter speakers from making truthful statements or present opportunities for government officials to selectively prosecute speakers according to their points of view. A majority of current justices have opined that “restricting false statements about philosophy, religion, history, the social sciences, the arts, and the like” would “present a grave and unacceptable danger of suppressing truthful speech.” 45 These areas are “broad,” in which the Constitution’s “breathing space” may extend to insulate all false statements from regulation. 46 The reasons offered by the Court for this broad protection include the often ambiguous determination of falsity with this type of speech, the potential value of even verifiably false speech in honing public issue debate, and, most markedly, the extraordinary potential that government prosecutors would abuse the power to restrict false speech to achieve political ends. 47
Governments at all levels may, and routinely do, impose liability for false statements of fact within realms of regulation in which the Court has tolerated content-based restrictions without comment or otherwise presumed the restrictions to be consistent with the free speech guarantee. No overarching theory distinguishes these realms from others in which restrictions provoke free speech inquiry. 48 These realms of regulation involve securities, antitrust, labor and employment, copyright, trademark, evidence, professional regulation, contracts, and many torts that may occur by means of speech. 49 Although some types of liability may require proof of knowledge of falsity by the speaker, other types of liability, such as for breach of contract or for faulty instructions on products or in connection with services or failure to warn may be imposed at the lower fault level of negligence or even without fault on the part of the speaker at all. 50
Although, for the most past, these laws impose liability for false statements made in one-on-one transactions, or in speech transactions within the realm, such as corporate board or labor elections, in some instances they may reach to restrict misinformation disseminated publicly. One such reach is professional discipline of lawyers who make false statements about public issues in court, or perhaps outside it. Several courts imposed sanctions against lawyers who repeatedly made, and perpetuated, unsubstantiated claims of fraud in the 2020 presidential election. Another court suspended former President Trump’s lawyer’s license, reasoning that such intentionally false statements “damage the proper functioning of a free society” and “tarnish the reputation of the entire legal profession.” 51 Such applications of professional rules of conduct to restrict public issue speech will focus the attention of courts on free speech issues that had not before appeared salient and will provoke the inevitable contextual determination of whether the fit between the speech restriction and the harm it is designed to avoid is sufficiently tight to comply with constitutional limits. 52
Regulations aimed at truth in advertising may restrict false speech or, within the confines of the Court’s expanding interpretation of unconstitutional compelled speech, may require vendors to provide additional information in their advertisements or product labels. Many of these laws exist at the federal and state levels, often enforced by specialized agencies. At the federal level, the Federal Trade Commission (FTC) is the agency primarily charged with prohibiting unfair competition and deceptive acts or practices. 53 An advertisement is deceptive if is contains a material misrepresentation or omission likely to mislead consumers to act to their detriment. 54 To impose liability, the FTC need not prove that an advertiser intended to convey misleading information or that consumers were actually injured by it. The information conveyed need not be verifiably false. Advertisers must affirmatively substantiate their claims and must present accurate endorsements, testimonials, product demonstrations, and comparisons. Liability attaches to express or implied claims. Enforcement often begins with warning letters to sellers making deceptive or scientifically unsupported claims. 55 The FTC may escalate remedies, seeking orders such as that the vendor cease and desist publishing the advertisements, issue corrective advertising or other education, or excise a trade name. If the vendor fails to comply, the FTC may seek civil penalties or a court order, and if vendors defy court orders, seek to hold them in contempt. The FTC polices deceptive advertising across a wide range of commercial activities and focuses its efforts particularly on health-related claims. A new statute granted the FTC additional authority and resources to target deceptive practices related to the COVID-19 pandemic. 56
The Food and Drug Administration (FDA) also has authority to police certain types of deceptive product and service advertising and is primarily responsible for regulating food labeling. 57 Only FDA-approved nutrition and health claims may appear on food labels. The FDA must approve drugs before they are marketed in the United States. To obtain approval, vendors must provide evidence that their products will perform as they will claim. These pre-approval processes give the FDA front-end authority over the truth and completeness of what vendors may say on their labels and in their advertising.
An unchallenged realm of regulation for decades, the FDA’s authority to regulate vendors’ speech is now contracting as the Court’s interpretation of the constitutional protection for commercial speech expands and lower courts follow this precedent. 58 The FDA’s order that cigarette packages include graphic warning labels has been delayed a decade due to ongoing litigation over whether it violates vendors’ free speech rights. 59 Laws imposed by states and cities requiring information disclosure by vendors have been challenged successfully. In one recent case, a court held that a requirement that sugary drink vendors warn consumers that their products “contribute[] to obesity, diabetes, and tooth decay” likely violated the free speech guarantee. 60
State agencies also enforce prohibitions against false statements and requirements that product contain accurate branding. Many states around the country have adopted the Uniform Deceptive Trade Practices Act (UDTPA), which bars misrepresentation, product disparagement, bait-and-switch advertising, and other forms of deceptive advertising. In some states, consumers, in addition to the government, may enforce false advertising laws. 61
Fraud is intentionally deceptive speech calculated to deprive another of something of value. The justices agree that when the government proves these elements to be present, it may impose civil or criminal liability on the speaker. Federal and state laws criminalize fraudulent statements for material gain in a variety of specific contexts, such as insurance fraud, securities fraud, and mail and wire fraud, including use of the internet, bank fraud, and credit card fraud. 62 Many federal and state online crimes may hinge on deception or false statements. These include identify theft, spoofing and phishing, predation, online impersonation, 63 social network fraud, 64 trafficking in passwords, 65 cyber extortion, 66 and cyberbullying. 67 Federal laws impose or authorize civil liability in a range of contexts. 68 State tort laws generally create causes of action for fraud and more specifically authorize civil suits for fraud in specific contexts. 69
When the elements of an intent to defraud and an individual subject to, or at risk of being deprived of property or a legal right, are present, fraud laws may be applied to impose liability on speakers for spreading misinformation through modern media that causes public harms beyond the target of the misinformation. In one recent prosecution the U.S. Department of Justice (“DOJ”) charged a woman with wire fraud and making false statements related to health care matters for selling fake COVID-19 immunization pellets and falsifying vaccination cards. 70 The woman’s offenses, according to the prosecutor, included “defraud[ing] and endanger[ing] the public by preying on fears and spreading misinformation about FDA-authorized vaccinations.” 71
Fraud as framed by law, however, may be an insufficient basis to impose liability in the common circumstances where web hosts intentionally post sensationally false claims for the purpose of attracting viewers, which translate to advertising revenue. The early example of the apolitical Macedonian teenagers who created websites aimed at Trump supporters to generate income from penny-per-click advertising to fund lavish (for them) lifestyles illustrates the phenomenon. 72 The teens knew that much of the content they published was false, but they did not defraud their viewers of something of value. The injury was to the democratic process, and the public generally. The possibility of restricting this type of speech is intriguing, but whether the Constitution permits laws to be interpreted, applied, or created to restrict publications of this type, involving public issues, is far from clear. 73
State defamation laws and laws establishing related speech torts originated in common law but are now often incorporated into statutes. Defamation laws permit individuals to bring lawsuits against speakers who transmit falsehoods that injure the reputation of an individual, organization, or corporation. 74 False-light invasion of privacy and intentional inflection of emotional distress causes of action may impose liability for false statements that cause the harms they protect against. 75 As set out above, heightened fault showings apply when the speech involves public officials, public figures, or issues of public concern. 76 Additionally, many states have so-called anti-SLAPP statutes. SLAPP stands for “Strategic Lawsuit Against Public Participation” and the purpose of the statutes is to add a layer of protection for publishers of public issue speech. The statutes generally permit speakers charged with defamation in the context of speech of public concern to file motions to dismiss lawsuits that lack sufficient legal grounding at an early stage of the proceedings, limiting the cost of defending against such claims. 77
The Constitution and state statutes present high hurdles to succeeding in a defamation action involving public issue speech. 78 The expense of litigation and practical and strategic concerns may also deter individuals from bringing suit. 79 Despite these barriers, some individuals can prevail. Recently and after many years of litigation, parents subject to a barrage of claims by Alex Jones, founder of the InfoWars website, that they had fabricated their children’s deaths in the Sandy Hook Elementary School shooting, won default judgments. 80 Much dangerous misinformation, including flat-out lies, circulate around elections, yet this is the realm where the constitutional standards seem almost impossible to meet. 81 Lawsuits alleging defamation by means of baseless claims of voting fraud, brought by individual voters and election technology companies, are pending and, if successful, will present an innovative means of restricting election-related disinformation. 82
Laws that criminalize defamation continue to exist in a number of states. 83 Even with the heightened “actual malice” fault threshold in place, these laws exist in tension with free speech values, since government officials may use them to protect themselves from criticism. 84 Courts in a number of states have held the laws to violate the Constitution, and many state legislatures have eliminated or narrowed their application. 85
Members of the Court agree that many laws criminalizing intentionally false statements to or about the government comport with the free speech guarantee because they are narrowed in scope to circumstances where harms to government interests or processes are likely to occur. 86 False reporting statutes criminalize false claims of crimes, catastrophes, or other emergencies—so-called hoaxes—or threats to commit acts of terrorism or other crimes. 87 These laws vary across the states, but generally narrow their application to intentionally false speech that does, or is likely to, cause public alarm or divert response resources. 88 Federal and state officials used these types of laws to prosecute individuals for perpetrating some of the thousands of anthrax scares after the Twin Towers attack of 2001. 89 These applications generally involved discrete, verifiably false claims through words or conduct occurring outside the context of public issue speech, and so are like the false statements involving government interests or processes or “falsely shouting fire in a theatre and causing a panic” that the Court has long presumed to fall outside the protection of the free speech guarantee. 90
Increasingly, however, enforcement of false reporting statutes is based on speech over the internet or on social media. 91 A number of individuals have been convicted for false posts threatening to infect people with COVID-19. 92 These applications involve false claims that are broadcast publicly and, because of the public significance and politicization of the pandemic, arguably occur within speech of public concern, and so raise more complex constitutional questions than the other false statement prohibitions. Courts will resolve the constitutionality of the statutes and applications in the context of enforcement actions. The statutes must identify harmful false speech without sweeping so broadly that prosecutors may apply the laws “subtly but selectively to speakers that the Government does not like.” 93 Applications must also overcome the argument that “more speech” quickly disseminated over the internet or social media is the proper remedy rather than criminalizing the false claims. 94
The Constitution requires “breathing space” protection for false statements that occur within election speech. So, the government’s ability to regulate to prevent the spread of election-related misinformation is severely limited. The federal government criminalizes false claims about election procedures, as do states. 95 Because false statements about the mechanics of voting can be verified and do not have instrumental value in protecting speech about the subject matter of the election, these types of laws should comport with the free speech guarantee. 96 Although enforcement of them aims only at a narrow class of election-related misinformation, it is a type that is particularly dangerous to the effective operation of democratic processes.
Recently, the federal DOJ charged a powerful social media influencer with conspiring to violate voters’ constitutional rights by using social media platforms, including Twitter, to disseminate fraudulent messages designed to persuade Democratic voters in the 2020 presidential election that they could cast their votes by means of text message or social media. 97 “There is no place in public discourse for lies and misinformation to defraud citizens of their right to vote,” the DOJ said in its press release. 98 In another prosecution related to the 2020 election, federal and state authorities charged two men with voter intimidation and computer crime for targeting thousands of robocalls to minority voters falsely telling them that information on their mail-in ballots would be used to track them by law enforcement, credit cards companies, and health authorities imposing mandatory vaccine requirements. 99 A federal judge rejected the defendants’ free speech defence and ordered them to make corrective calls days prior to the election. 100
The federal government does not regulate false statements about the subject matter of elections, but many states do so. 101 These laws impose liability for a number of different types of false claims. Many of them do not explicitly or implicitly require an “actual malice” level of fault by the speaker, which would seem to be a constitutional threshold with respect to any laws of this type. And, although the purpose of protecting election integrity may be compelling, the narrow tailoring requirement is very difficult to meet. 102 So, despite the great number of such state laws, many are likely not enforceable. False statements made by judges or candidates in judicial elections are an exception that the Constitution may permit states to regulate more extensively through ethics rules and professional discipline because of the special role of judges in the justice process. 103
A number of states have false statement laws that explicitly apply to online actions and are broad enough that they could be applied to election deception perpetrated over the internet. 104 Some state laws specifically target online election deception. For example, California law prohibits “political cyber-fraud,” which includes intentionally causing visitors to believe advocacy or information posted on a political website represents the views of candidate or someone in favor or opposed to a ballot measure, when it does not. 105 Texas law, and California law through 2022, prohibit distribution of election-related “deepfakes,” which use artificial intelligence to create highly realistic videos that appear to portray real people when they do not. 106 Courts will review the terms of the laws and applications in enforcement actions to determine whether they meet the Constitution’s strict narrow-tailoring requirements in the highly protected realm of election speech. Other laws implement options to counter the harm of political and other types of deepfakes without restricting them and so more likely are consistent with the free speech guarantee. A new California law will require disclaimers on certain election-related deepfakes, as set out below. Federal law calls for a report on the foreign weaponization of deepfakes, requires defense agencies to notify Congress of foreign deepfake-disinformation activities targeting U.S. elections, and establishes a “Deepfakes Prize” competition to encourage the research or commercialization of deepfake-detection technologies. 107
Congress and state governments regulate election funding. 108 Consistent with the Court’s interpretation of the free speech guarantee, governments may impose some limits on contributions by domestic individuals or entities to campaigns and on spending for communications coordinated with campaigns 109 but may not limit their spending on so-called “independent expenditures,” which advocate expressly for the election or defeat of a candidate but are not coordinated with the campaign. 110 Since independent expenditures are a significant source of false election advertising, the inability to limit such expenditures cuts off one avenue for indirectly curbing this source of misinformation.
In any event, federal law prohibits participation by foreign nationals in U.S. elections more broadly, prohibiting both contributions to candidates and expenditures to advocate their election or defeat directly. 111 It does not, however, currently prohibit expenditures by foreign nationals on more general types of speech, such as the misinformation broadcast in social media advertisements purchased by Russian operatives during the 2016 presidential election. Federal law also does not prohibit independent expenditures by domestic subsidiaries of foreign corporations in candidate elections or by foreign nationals related to state elections that involve issues rather than candidates, such as ballot initiatives. 112 Some state laws fill these gaps. 113 A federal law that requires foreign agents to register and disclose certain types of communications aimed at influencing U.S. politics may cover some scope of foreign election-related communications not currently prohibited by federal election law. 114
With speech and spending restrictions largely unavailable as options, transparency requirements are the primary means by which federal and state regulators address the harms of misinformation in election speech. Disclosures require political advertising funders to submit reports to regulators of their expenditures and allow voters to “follow the money” to understand how much money political groups are spending and for whom. 115 Disclaimers appear on political advertising and identify the sponsor. They allow voters to evaluate the content of the messages at the moment they receive them, according to the credibility of the speakers and likely biases. The Court has continued to uphold transparency regulations against free speech challenges, accepting the government’s purpose to “enable[] the electorate to make informed decisions and give proper weight to different speakers and messages” and noting that they are a preferable alternate means to restricting speech. 116
Gaps in transparency laws limit their effectiveness. Failure by law to require disclosure by political nonprofits creates a loophole whereby large amounts of so-called “dark money,” because its source is secret, can fund election speech without accountability. 117 Federal law does not require online entities that distribute or publish political advertising to retain or publish records of them, which hinders enforcement of disclosure requirements against sponsors. 118 The disclaimer rules exclude the significant amount of election-related speech posted on the internet for free, which may be disinformation and refer specifically to candidates. 119
The specifics of state campaign finance rules vary, but they generally include campaign contribution and advertising regulations similar to the types in federal law. While the federal government has lagged, a number of states have enacted disclosure and disclaimer rules specific to the internet. 120 A number of these place the disclosure responsibility on the platforms that sell space for the advertising and require that the records be publicly available online. 121 Two California laws require disclaimers on election-related use of the new technologies of deepfakes and “bots.” 122
Traditional media refers to any form of mass communication available before the advent of digital media. This includes print media such as books, newspapers, and magazines, media that travels over airwaves, such as radio and broadcast television, and the more modern forms of cable and satellite television.
Under common law principles, others in addition to the original speaker who rebroadcast falsehoods may be held liable according to their degree of editorial control over the material. Publishers, such as newspapers, television stations, and radio broadcasters, exercise substantial editorial control over the content of their publications and are deemed responsible for its content. 123 Distributors, such as book stores or libraries, may be held liable only if they knew or should have known about the content. 124 Common carriers, such as the post office or telephone companies, are passive conduits that exercise no editorial control over the communications they transmit, and so are not subject to republication liability even if they are made aware of the content of the communications they transmit. 125 While republication liability expands the potential scope of defamation recovery, state laws impose various types of retraction requirements and limits to liability. 126 Traditional media outlets that publish their content over the internet may acquire the immunity from republication liability provided by section 230 of the Communications Decency Act, described below, in locations within their sites where they host content created by others, such as viewer comment boards.
The Court has long interpreted the Constitution to prohibit the government from imposing content controls on print content, even if it may risk grave public danger to publish it. 127 No federal agency grants licenses to print media, or journalists, to operate, and no justification for doing so, such as allocation of scarce resources, exists. Print media entities receive the same scope of constitutional protection, and are subject to the same scope of liability, for the content of the speech they choose to broadcast as individual speakers. Public opinion and private lawsuits are the means of encouraging them to avoid publishing false speech. 128
The Federal Communications Commission (FCC) regulates telecommunications media—telephone, telegraph, radio, television and broadband internet providers—to varying degrees. The FCC grants licenses to local radio and television stations. In exchange for the license, and according to Congress’ mandate, the FCC requires licensees to operate in ways that serve the “public convenience, interest or necessity,” which, its website now states, “is best served by permitting free expression of views.” 129
In furtherance of this philosophy, the FCC’s exercise of content control over licensees has declined over time. Still, its regulations prohibit licensees from broadcasting false content in two situations—so-called “broadcast hoaxes” 130 and “news distortion.” 131 The FCC rarely enforces these rules, viewing them as applicable only in extreme circumstances. 132 The agency quickly and firmly rejected a recent petition that asked it to advise licensees to qualify COVID-19 misinformation in statements by former President Trump and from other sources with a disclaimer that the information was false or scientifically suspect. 133 It is not “arbiter of truth in journalism,” the FCC said. 134
The FCC imposes certain limited equal-opportunity requirements on radio and television entities in connection with appearances by candidates in elections 135 The FCC also enforces statutory transparency requirements on the entities it regulates, which supplement the political advertising requirements impose directly on sponsors. The regulated entities must also create disclosure files with information about election advertising and upload them to the internet. 136
With little or no government content control in the background, self-regulation has been and continues to be the primary means for controlling the content delivered by traditional media outlets. At one point, the majority of broadcast radio and television licensees followed industry codes of conduct. 137 Now, the separate standards and practices departments of the major networks, both broadcast and cable, evaluate and approve program content. 138 Journalistic codes of ethics exist with respect to news reporting. 139 Individual news organizations have similar codes of journalistic ethics. Media entities may discipline journalists working within their enterprises who violate the standards, creating public scandals that reinforce norms of ethical reporting and impact internal review procedures to avoid recurrence. 140 Contributors within a news organization may resign to protest violations of journalistic ethics. 141 News organizations increasingly provide venues, such as online forums, for readers or viewers to comment on the accuracy or fairness of the reporting they publish, and peer-reviewed academic publications impose some accountability. 142
Section 230 of the Communications Decency Act, enacted as part of a larger overhaul of telecommunications law in 1996, provides special legal protection to internet service providers from liability for content created by others. The core purposes of section 230 were “to promote the continued development of the internet and other interactive computer services” and “to preserve the vibrant and competitive free market that presently exist[ed] for the internet and other interactive computer services, unfettered by Federal or State regulation.” 143
Section 230 has two main provisions. Section (c)(1) states, “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” 144 This provision shields internet intermediaries from liability for hosting or otherwise facilitating the transmission of false statements, or other communications with content that may form the basis for a legal claim, so long as the content was created by others. 145 This section responded specifically to a state court ruling that had held an early online service that hosted a variety of content produced by other providers liable for defamatory content posted by a user on a message board. 146
Section (c)(2) states that providers are not liable for “any action voluntarily taken in good faith to restrict access to or availability of “content the providers determine to be objectionable “whether or not such material is constitutionally protected.” 147 This second provision enables internet hosts to edit and otherwise restrict the content they transmit, which includes removing false content or forbidding purveyors of false information from using their services. According to its sponsor’s statement, this section was designed to encourage “computer Good Samaritans” to assist computer service customers, particularly parents, to control what online content comes into their homes and what their children see. 148
Courts have interpreted the section 230 to afford a broad scope of protection to entities that qualify as providers or users of interactive computer services and that do not qualify as content providers or creators. 149 An influential decision shortly after section 230 was enacted held it to shield computer service providers from a broad range of claims based on speech content, not only defamation, and in instances where the provider would otherwise face liability as a distributor because it received notice about the harmful content and failed to take it down. 150 Section 230 protects computer service users as well as providers, such as persons who forward or repost content, so long as they did not exercise significant control over the content, 151 and protects social media platforms from liability for the user content they host. 152 Section 230 shields service providers from liability as publishers and does not extend to other types of claims. 153 It also requires that their content modification actions be made in “good faith” to achieve the purposes set out in the statute, and not for anticompetitive purposes. 154
Congress intended section 230 to encourage the free flow of information over the internet and it has done so. In fact, as noted by the Court, the internet, and social media, have become more important places for public communication than government-owned parks or streets where citizens traditionally gathered to speak. 155 Yet in those physical spaces, the Constitution places the government in the role of public trustee, required to grant access to all speakers and forbidden to discriminate among them according to points of view. 156 Similarly, Congress envisioned a public trustee role for internet platforms, but charged them with the responsibility to engage in viewpoint discrimination among the speakers they host, rather than ignore the content of their speech and treat them equally.
This contradiction has not gone unnoticed. Section 230’s advocates laud the range of types of speech that would not be available without its protections, and the efforts and accomplishments of internet hosts in screening and removing dangerous content from their platforms. 157 Critics contend that self-interest most often prevails over public responsibility in content moderation decisions, that victims of online abuse cannot seek redress, and that platforms have too much freedom to selectively silence point of view. 158 Reflection at the federal and state levels is ongoing, about whether regulatory changes are appropriate and what they should be, and speculation about whether the Court would find regulations to be consistent with the free speech guarantee continues.
The FCC does not currently assert authority to regulate the content of speech on the internet. Congress could choose to expand the FCC’s jurisdiction over internet service providers, or otherwise to implement by statute a means to oversee the terms and technologies by which internet service providers host and moderate content, subject to judicial review for consistency with constitutional limits. 159
As noted above, some states impose some disclaimer and disclosure requirements on internet platforms with respect to election speech. Commentators propose that the federal government and states enact many more disclaimer and disclosure requirements applicable to those communicating and the platforms hosting the communications. 160 The Honest Ads Act, pending in Congress, would update internet disclaimer and disclosure rules to impose the same types of requirements currently imposed on radio and television platforms with respect to political advertising. 161 Many proposals to modify the scope of section 230 immunity for internet service providers to require or incentivize them to exercise greater or more responsible control over the content they post have been or are pending in Congress, but none has yet been enacted into law. 162
Data privacy laws address the spread of disinformation indirectly by limiting the ability of those who want to disseminate disinformation to collect personal data that allows them to micro-target recipients. No general national data privacy law exists. A mix of federal laws protects specific types of data in certain circumstances. 163 Three states have general data privacy laws that allow consumers to opt out of data-gathering. The California Consumer Privacy Act (CCPA) provides the greatest protection, including a “global opt-out” and a limited private right of action. 164 The application of all of these state laws is narrower than the European General Data Protection Regulation (GDPR), which extends beyond consumer protection. 165
Section 230 leaves decisions about how to locate and what to do about misinformation on online platforms to the service providers. Online operators maintain policies that prohibit users from posting certain content, including misinformation. 166 The policies are in the discretion of the providers and may change over time. 167 The platforms all have different policies, and many different factors enter into determinations of what constitutes misinformation that violates the policies, so it is difficult to generalize about the substance or features of misinformation platforms will choose to identify and modify. 168
Some common procedures exist. Operators rely on several sources to identify false content. These include users, hired human content moderators, and automated systems, also known as artificial intelligence technologies (AI). 169 None of these are simple to deploy. For example, using AI to screen for misinformation is attractive to a massive social media platform like Facebook because of the scale of the project. But AI, although automated, still needs to learn and so cannot react quickly to detect misinformation related to sudden events. 170 And humans need to perform the second-stage, fact-check, review in most cases to determine what to do.
The use of fact-checking organizations has proliferated in the past five years with more social media organizations relying on their services to evaluate what constitutes misinformation on their platforms. 171 Facebook provides a window into the size of the fact-checking undertaking. As of 2017, Facebook was reviewing 100 million posts per month, 172 and by 2020, it employed more than 80 fact-checking organizations across the globe that reviewed content in over sixty languages. 173 Once identified, operators respond to misinformation on their platforms in a number of ways. 174 These include adjusting algorithms to prioritize accurate content, labeling false content, demoting false content so it appears less frequently in searches, removing false content, penalizing users who post misinformation repeatedly by removing certain privileges, 175 or banning them from the platform, temporarily, or permanently. 176
These efforts exhibit a mix of self-interest and responses to various public pressures. As one example, and after public attention had been directed to the spread of COVID-10 misinformation on their platforms, major operators issued a joint statement pledging to respond more forcefully. 177 And, indeed, all the major social media companies took specific actions to identify and remove misinformation and the amplify correct content. 178 Still, more reports of how pandemic misinformation spread rampantly on their platforms appeared. 179 The President and other government officials criticized their efforts as inadequate. 180 Congress called platform heads for hearings. 181 Private advocates staged protests. 182 The platforms disputed the critiques and posted reports on their sites of how much misinformation they had identified and removed. 183 And so the back-and-forth continues, with platforms purportedly open to some types of regulation, but Congress uncertain and ununified about which way to go. 184
Promoting diversity of points of view in media is a means to counter the spread of misinformation of encouraging the broadcast of more speech, including counter-speech, rather than restricting false statements.
The Constitution protects print media entities like individual speakers and so the government may not regulate the content they choose to publish, either to prohibit falsity directly or to promote accurate communication indirectly, by imposing diversity requirements with respect to content or ownership. Consistent with the Constitution, federal government policy for over a century promoted press diversity by means of subsidization, encouraging the growth of newspapers and magazines by providing for extremely low rates for transport by the government-owned post office. 185
The FCC continues to implement some content and ownership policies to encourage localism and diversity in radio and television broadcasting. 186 Taxpayer support for public broadcasting, albeit diminished over the years, is another way the government promotes diversity in broadcasting and access to it. 187 Cable television operates under a mix of federal and local regulation to promote diversity and preserve access to local content. The FCC applies certain access requirements to cable system operators. 188 Although the FCC has the authority to impose vertical and horizontal ownership caps, none are currently in place. 189 Local governments enter into contracts with particular cable operators that set terms and conditions in exchange for the right to use public infrastructure. These contracts usually require cable operators to carry certain broad categories of programming, including public-access, educational, and government channels. 190 The FCC also applies access rules to satellite television operators. 191
“Net neutrality” is the internet access equivalent, requiring that the corporations that provide internet service to consumers provide the same service to all of them, regardless of how much they pay. 192 The concept has been controversial in the United States, and the FCC’s application of it, confusing. Although the FCC implemented it for a few years, it no longer does so. 193 President Biden has issued an executive order urging the FCC to restore the net neutrality rules. 194 How the Court would classify a net neutrality rule for free speech analysis and whether it would find it consistent with the Constitution is unclear. 195 California enacted a net neutrality law, as did Vermont, which has put its law on hold pending resolution of a lawsuit by internet providers challenging the legality of the California law. 196
The federal DOJ and FTC and similar state agencies regulate vertical and horizontal anticompetitive actions among media entities under general antitrust laws. Early on, the Court rejected a free speech challenge to antitrust enforcement with respect to a wire service selling news to print media, reasoning that “a free press is a condition of a free society” and the Constitution’s guarantee “rests on the assumption that the widest possible dissemination of information from diverse and antagonistic sources is essential to the welfare of the public.” 197 The federal government has a long history of applying antitrust laws in the newspaper sector and, with the rise of competing new media and the incentive to consolidate, continues to target it for enforcement. 198
For many years, the FCC enforced licensing rules that prohibited common ownership of newspapers, radio stations, and television stations within the same area, noting “powerful reasons” for the government to prohibit combinations that “restrain trade in news and views.” 199 In 2017, it eliminated most of these rules, reasoning that changes in the media market sufficiently accomplished the diversity goal and thereby rendered ownership restrictions “obsolete.” 200 The agency continues, however, to review transfers of media interests to ensure that they are consistent with its broad “public interest” mandate. 201 The DOJ website lists over ten selected investigations and cases in the last decade, 202 and practitioners speculate that the FCC’s retreat may signal increased antitrust enforcement by the DOJ, FTC, and state officials against media entities. 203
Federal and state authorities are increasingly turning to antitrust enforcement to limit the power of the broadband providers and the search engines and social media platforms that control the flow and content of information on the internet. A recent executive order urged the FCC to more closely regulate the rates charged by broadband providers. 204 Antitrust inquiries into the behavior of internet platforms aim to incentivize them to exercise more responsible management of the content of the speech they host. 205 Platforms are facing charges filed by federal and state agencies 206 and private lawsuits for antitrust violations. 207
Most proposals to combat the modern misinformation problem call for improving media literacy, meaning the ability to impart, receive, and effectively evaluate information by means of all forms of media. 208 Media literacy is broadly defined as the ability to access, analyze, evaluate, create, and act using all forms of communication. 209
State laws addressing media literacy education began to appear in 2008. 210 A 2016 study by a group of Stanford researchers found American students had difficulty identifying the credibility of online information—like information found on social media platforms. 211 This study, along with the experiences of the 2016 presidential election, motivated more state legislatures to enact laws so that four years later fourteen states were “on the[] way to establishing media literacy curriculum as a priority in K-12 schools.” 212
No laws specifically mandating or providing support for media literacy education exist at the federal level. A Senate bill, The Digital Citizenship and Media Literacy Act, would provide 20 million dollars of federal funding for digital citizenship and media literacy in K through 12 education. 213 Training in digital citizenship education adds lessons in the safe, responsible, and ethical use of communication technologies to media literacy education. 214
Public education is a soft-law means by which governments may influence the behavior of recipients of misinformation. Outreach in response to the COVID-19 provides an example of government rumor control and counter-messaging efforts. Federal efforts include a COVID-19 Disinformation Toolkit created by the Cybersecurity and Infrastructure Agency (“CISA”), 215 a public education campaign on vaccines on the Department of Health and Human Services’ website—”We Can Do This,” 216 and a rumor-control webpage in question-and-answer format hosted by the Federal Emergency Management Agency (FEMA). 217
State and local agencies provide this type of citizen information as well. The California Department of Public Health established a trust and safety team that encourages anyone who sees or hears of any vaccine-related rumors across social media or in their communities to report them via a dedicated “Rumors” email address. 218 Sacramento County hosts a webpage in the same rumor/fact format as FEMA. 219 Additionally, the federal government and some states have sought out and paid social media personalities or “influencers” on Twitch, YouTube, and TikTok to reach their many followers to combat vaccine misinformation. 220
Efforts by nonprofit organizations and other nongovernmental entities complement the government efforts and aim more broadly to educate speech recipients about the prevalence and danger of misinformation and prime them to accept correction. These efforts include online games that teach users how to distinguish true and false claims, “pre-bunk” (rather than debunk) common false claims, or provide “accuracy nudges” to encourage users to reflect upon the accuracy of claims before accepting them. 221
Fact-checking the news is an exercise that the government could not possibly do with credibility. 222 Fact-checking sites have grown in response to the internet, focused primarily on verifying or correcting claims published on it and relying on the access to information it provides to make the exercise viable. 223 As of a few years ago, most U.S. fact-checkers were professional journalists. Provoked by the focus on misinformation and fueled by the “fact-checking” philanthropy of internet providers and nonprofits, more sites have sprung up, so that fact-checking is now run by a mix of news organizations and independent entities. 224 Fact-checking originally focused on elections, with large national sites and columns such as FactCheck.org and PolitiFact focused on national politics and state and local fact-checkers, most run out of news organizations, covering those races. 225 Fact-checking sites now cover a range of topics. Lists of fact-checking sites, with links, are available on the internet. 226
With the rise of fact-checking as an antidote to misinformation, has come research on its efficacy. 227 Efforts to improve its accuracy and effectiveness in convincing recipients to reject misinformation continue. 228 The bottom line, at this point, seems to be that “it works, within limits” 229 and so it remains an important, if insufficient, tool to stem the spread of misinformation.
This Report provides an overview of the U.S. free speech guarantee, as interpreted by the Court, the limits it imposes on regulations to stem the spread of disinformation, and the various forms of regulation that exist within its framework. U.S. free speech doctrine and fake news regulation are both complex and evolving. This Report provides a sketch and snapshot of this important intersection to aid understanding and assist comparison with other nations’ efforts to address the extraordinarily pressing global problem of the spread of misinformation.
Thanks to Arash Aalem, James Silverthorn, and Matt Urban for research and cite-checking assistance.