Issue Area 1: Balancing Free Speech with Appropriate Restrictions at the National Level
A. New Issues, Old Laws
Despite the development of international human rights principles, States struggle to balance the right to freedom of expression with restrictions in accordance with legitimate interests that are recognized under international law. States have long used discrete laws dealing with defamation, elections, consumer protection, and financial fraud to address the harm done by false information. Even in exceptionally charged political environments, courts have been able to rely on these frameworks to adjudicate election disputes.
The United States, for example, historically has employed a very broad interpretation of free speech. The First Amendment to the U.S. Constitution states unequivocally that “Congress shall make no law … abridging freedom of speech,” a right that has grown to encompass many forms of citizen speech. However, defamatory statements, which are defined as “false statements of fact about a person,” incitement, fraud, obscenity, child pornography, fighting words, and threats are not protected speech under the First Amendment. In defamation cases, the party alleging defamation must “demonstrate that the speaker acted with a certain level of intent … or to prove certain injuries.” Certain federal laws regarding elections also prohibit false statements about voter eligibility and “fraudulent misrepresentation of authority to act for a federal political candidate.”
Dominion Voting Systems, Inc. v. Fox News Network, Fox Corporation illustrates how U.S. courts have grappled with issues of free speech and defamation in the context of elections. Prior to the 2020 U.S. presidential election, analysts, including New York-based Fox News, predicted that early vote counting would favor the incumbent, Donald Trump, and that later counting would favor his opponent, Joseph Biden. This was indeed how the election proceeded, with Biden eventually securing the presidency. Infuriated by Fox News’ projections and a potential loss of the presidency, Trump and his supporters adopted a narrative that widespread fraud had tainted the election. As Fox News’ viewership dropped, the network suddenly began to amplify Trump’s false narrative, presumably in hopes of boosting its ratings. Shortly after the election, Fox News began stating publicly that Dominion Voting Systems, Inc., a private supplier of election and voting technology that operated many of the voting machines used during the election, had used those machines to “rig” the election in favor of Biden.
Dominion, in turn, alleged that Fox News’ statements were false and defamatory and sought punitive and economic damages. Fox News asserted various defenses, chief among them that the network’s reporting of newsworthy events was protected under the First Amendment, statements were not published with actual malice (i.e., intended to cause harm, which Fox stated that Dominion did not show with “clear and convincing evidence”), and Dominion did not suffer any damages. In fact, Fox News argued specifically that a reasonable viewer would understand the network was simply “fulfilling its journalistic duty to present newsworthy allegations made by others.” The network also asserted that New York case law had established that, when the press repeated allegations later found to be false, a reasonable viewer would understand the allegations as “claims” rather than “facts.”
In rejecting these lines of defense from Fox News, the Superior Court of Delaware found that the statements were defamatory per se because the network claimed that Dominion committed election fraud, among other key issues, and that “[n]o First Amendment protection enfolds false charges of criminal behavior.” Relying on longstanding jurisprudence around defamation in the State of New York, the Court was able to resolve this significant case. Its finding of defamatory statements per se and no defenses of freedom of speech led Dominion and Fox News to settle the case out of court for over USD $750 million.
B. National Legal Frameworks
A prescriptive legal framework directed at controlling the flow of information can often generate more challenges before a court. In some African jurisdictions (such as Uganda, Zambia, and Zimbabwe), laws dating to European colonization criminalized the spread of loosely defined “false information” on issues of public interest and have been found to be both unconstitutional and unjustified in modern democratic societies. In the last decade, the number of laws prohibiting “false news” has also risen, most notably in Asia and across the Pacific. Though many of these laws remain in effect, they ultimately fail the “necessity and legitimacy” tests set out under the ICCPR.
Countries navigate the problem of disinformation in various ways. In Brazil, the Brazilian Criminal Code contains three provisions dealing with an attack on a person’s “honor,” but no current law specifically addresses disinformation. As of the writing of this paper, however, Brazilian Congressional Bill No. 2630, the Law on Freedom, Responsibility and Transparency on the Internet, is pending before Congress, already having been approved by the Senate. Although Nepal does not have a specific law addressing information integrity, the Libel and Slander Act, 2016, grants people the right “generally to maintain their prestige, honor and dignity;” the Electronic Transactions Act, 2063 (2008) prohibits the publication of material in electronic form “which may be contrary to the public morality or decent behavior or any types of materials which may spread hate or jealousy against anyone or which may jeopardize the harmonious relations subsisting among the peoples of various castes, tribes and communities;” and the Election Commission Nepal (ECN) Code of Conduct prohibits the transmission of “disinformation, misinformation and hate speech in social networks.” In Kenya, several articles of the Kenyan Constitution address freedom of expression and its limitations, and certain sections of the Computer Misuse and Cybercrimes Act make the publication of false information a crime. In Nigeria, a section of the Criminal Code addresses the publication of false news, including in instances when the person disseminating the news does not know it is false. Furthermore, Nigeria’s proposed Hate Speech Bill (which, after sustained public backlash, did not pass) would have criminalized any actions a person took to stir up ethnic hatred, engage in ethnic harassment, or discriminate. Such broad language would have granted the government arbitrary power to clamp down on free speech, including “critical opinion, satire, public dialogue, and political commentary” – a serious problem, especially during election periods.
Supreme/apex courts often must determine whether newly introduced legislation regarding information around elections aligns with constitutional provisions by following their jurisdiction’s appropriate test for limiting free speech. Recently, supreme and constitutional courts globally have struck down a variety of “false information” laws, arguing that the provisions of those laws did not meet the tests of necessity and legitimate aim to restrict freedom of speech and freedom of the press during elections. In South Korea, the Constitutional Court was called on to rule on the constitutionality of a legal provision that prohibited candidates from publishing a column, comment, contribution, or writing on online media within 90 days of an election. In this case, 2016Hun-Ma90 (Case on Restricting Online Media from Publishing Columns, etc. Written by Candidates for Public Official Election), the Court acknowledged that, while the restriction was intended to avoid unfairness in online election news reporting and the circumvention of campaign rules, it was overly broad because it prohibited the online publication of information that might not necessarily be political speech or otherwise tied to an election campaign. The Court found that the provision placed an unconstitutional restriction on the complainant’s freedom of speech.
During Nepal’s 2022 elections, voters created an online campaign expressing disenchantment with mainstream politics and career politicians. A member of one political party filed a complaint against the campaign, causing the ECN to warn the campaigners to cease campaigning or face fines, imprisonment, or both. The ECN relied on a broad provision of its Election Code of Conduct (“false or incorrect statement”) and referred the case to the Cyber Bureau, requesting that the police take down the campaign’s web pages. Responding to this warning, Senior Advocate Dinesh Tripathi filed a writ petition at the Supreme Court of Nepal, Senior Advocate Dinesh Tripathi v. Election Commission of Nepal (#NoNotAgain Campaign), arguing that the warning to cease campaigning violated the campaigners’ rights. The Supreme Court ordered the ECN and Cyber Bureau not to take any actions against the campaigners, stating that the campaign was an example of freedom of thought and expression. On November 6, 2022, the Supreme Court issued an interlocutory interim order against the decision of the ECN to refer the case to the Cyber Bureau until the final decision. Despite a lack of time to adjudicate this case on the merits, the Supreme Court’s quick actions nevertheless ensured that free speech would not be undermined during the election campaign.
In France, the Constitutional Council reviewed French Law No. 2018-1202 on the “fight against the manipulation of information” prior to its adoption in 2018. Unlike many supreme/apex courts that previously struck down provisions of the law, the Council dismissed the allegations that certain provisions of the law were in breach of freedom of expression. The Council acknowledged the legislature's responsibility “to bring an end to the abuse of the right to exercise freedom of expression and communication which infringe on public order and the rights of others” and found that the wording of the law was necessary, suitable, and proportional to the legislature’s aim of fighting manipulated information. Indeed, an interlocutory proceeding under this law has a limited scope, wherein only "incorrect or misleading allegations or accusations which have the effect of altering the honesty of the upcoming elections” fall within its purview. The law excludes opinions, parodies, partial inaccuracies, or simple exaggerations and only allows for three cumulative conditions for spreading such allegations or accusations: “they must be artificial or computerized, deliberate and spread by mass distribution.”
However, the Council also found that, “given the consequences that proceedings may have the effect of stopping the spread of certain information content, the allegations or accusations in question can only justify such a measure if the incorrect or misleading nature is apparent, without infringing on the freedom of expression and communication. Likewise, for the risk of having an effect on the sincerity of elections, which must also be apparent.” In light of the legislation’s limited scope and its strict definition of what constitutes manipulated information, the Council upheld the law as constitutional.
In Switzerland, the legal framework does not provide for strict regulations relating to disinformation, but general principles and a strong body of jurisprudence guide election judges in their application of the laws. For instance, the judges of the Federal Supreme Court have adopted a narrow interpretation of disinformation and determined that a court’s duty to intervene in disinformation cases can only happen when the influence of private actors seriously hinders or prevents the voters’ process of forming an opinion (see text box). Such conduct can lead to the annulment of a vote. Case law led to the use of the following test, relying on four criteria: “1) Erroneous information must first be based on facts (objective). 2) The facts must then relate to an important circumstance of such a nature as to seriously mislead the elector. 3) The disclosure of erroneous facts must take place at a late stage of the campaign, at a time when rectification would no longer have any effect on the voter. 4) Finally, the judge must satisfy himself that the misleading influence exerted on the electorate is without doubt or at least appears highly probable.” These four criteria echo the strict criteria adopted in the French law on manipulation of information and the rulings of other countries preventing unreasonable restrictions of freedom of expression.
Guidance for Swiss Election Judges
“The case-law has made it clear that only inaccurate and essential facts, which are available to the administration alone and which are not called into question by public debate, are capable of distorting the free formation of the will of the electorate. Information that may be erroneous, but which is the subject of the public discussion before the vote, is not enough. Similarly, imprecise estimates by the administration, but qualified as such, do not distort the democratic debate.”
—Judge François Chaix, of the Swiss Federal Supreme Court
Case law has helped establish the various categories of unprotected speech. Some prominent examples include Brandenburg v. Ohio (incitement); Virginia v. Black (true threats); Chaplinsky v. New Hampshire (fighting words); Miller v. California (obscenity). See Defamation. | Constitution Annotated | Congress.gov | Library of Congress (outlining a more detailed definition of defamation); Unprotected Speech Synopsis. The Foundation for Individual Rights and Expression (thefire.org) (defining unprotected free speech); United States Courts (n.d.) What Does Free Speech Mean? | United States Courts (uscourts.gov); fighting words. Wex | US Law | LII / Legal Information Institute (cornell.edu) (detailing the extensive case law behind “fighting words”)
52 USC 30124: Fraudulent misrepresentation of campaign authority.
U.S. Dominion Inc., Dominion Voting Systems, Inc., and Dominion Voting Systems Corporation v. Fox News Network, LLC, C.A. No.: N21C-03-257 EMD (Del. Sup. Ct., 2023) and U.S. Dominion Inc., Dominion Voting Systems, Inc., and Dominion Voting Systems Corporation v. Fox Corporation, C.A. No.: N21C-11-082 EMD (Del. Sup. Ct., 2023) (the cases were merged).
Id. at Sec. I. FNN'S ELECTION COVERAGE AND RISING BRAND CONCERNS.
Dominion contended that: “i) [(Fox News Network)] FNN and [(Fox Corporation)] FC intentionally provided a platform for guests that FNN’s hosts knew would make false and defamatory statements of fact on the air; ii) FNN and FC, through FNN’s hosts, affirmed, endorsed, repeated, and agreed with those guests’ statements; and iii) FNN, with the participation of FC, republished those defamatory and false statements of fact on the air, FNN’s websites, FNN’s social media accounts, and FNN’s other digital platforms and subscription services.” Id. at p. 3.
Id. at p. 31.
Both parties are incorporated in Delaware, and Fox News has its headquarters in New York. See id. at Footnote 230, which states “[t]he Court made comments at the March 21–22 hearing that Delaware law may control on punitive damages. After a review of the case law, the Court agrees with the parties that New York Law applies to the issue of punitive damages.”
Under New York State law, Dominion needed to establish: 1) a false statement; 2) publication without privilege or authorization to a third party; 3) constituting fault as judged by the actual malice standard; and 4) that causes special harm or constitutes defamation per se, which includes accusations of a serious crime or business harm. See id. at p. 37.
Op. cit., UNHCR Special Rapporteur p. 11.
For instance, Vanuatu and Tonga adopted new criminal defamation laws in 2021, contrary to good practice which typically sees defamation as a civil matter. This appears to be part of broader backsliding around governments’ protection of human rights in the region. See Lee, K., and Natalegawa, A. (2021, June 11). Fake News Crackdowns Do Damage Across Southeast Asia During Pandemic. Center for Strategic & International Studies (CSIS). Public Media Alliance (2019, December 6). The rise of “fake news” laws across South East Asia. CIVICUS. (2021). State of Civil Society Report 2021.
The three provisions include 1) slander, or the false imputation of a crime to another person (Article 138); 2) defamation, or the imputation of something offensive to a person’s social reputation (Article 139); and 3) injury, or imputation of something offensive to someone’s dignity. Brazil Media Law Guide (n.d.) Defamation, Privacy and Data Protection Law in Brazil. Carter-Ruck.
The bill itself is quite controversial, with no guarantee of becoming law. Often referred to as the “Fake News Bill” by its supporters and the “Censorship Bill” by its opponents, it would require technology companies and social media platforms to be much more transparent with their users about their content recommendations and take on more responsibility for third-party content being displayed on their platforms. See Martins, L., and Spagnuolo, S. (2023, April 28). A General Review on the Brazilian Congress Bill Regarding Fake News. Núcleo.; Al Jazeera. (2023, May 2). “Brazil’s ‘fake news’ bill sparks outcry from tech giants.”
Libel and Slander Act, Preamble, 2016 (1959) (Nep.).
The Electronic Transactions Act, 2063 Sec. 47 (2008) (Nep.). Such illegal conduct is punishable by a fine or imprisonment or both. Critics of Nepal’s disinformation policies claim this provision does not do enough to combat disinformation. See also Shrestha, P. (2023, February 19). “No policy to counter disinformation.” The Kathmandu Post.
The Election Code of Conduct, Para. 4j (2022) (Nep.)
The Computer Misuse and Cybercrimes Act, Para. 22 (2018) (Ken.)
Criminal Code, Sec. 59 (1990). Under this provision, a person found guilty of disseminating false news is guilty of a misdemeanor, punishable by up to three years’ imprisonment. Not knowing the information is false is not a defense unless the accused can show that they took reasonable measures to verify the information before publishing.
A Bill for an Act to Provide for the Prohibition of Hate Speeches and for Other Related Matters. (Nig.). Under the proposed bill, any person could file a written complaint with the Independent National Commission for the Prohibition of Hate Speeches, which could lead to conciliation or a hearing (or a dismissal, if the Commission determined the complaint to be lacking). The Hate Speech Bill was eventually dropped after intense public backlash. See Opejobi, S. (2019, December 4). "We won’t pass hate speech bill – Senate President, Ahmed Lawan." Daily Post.
Amnesty International (2019, December 4). Nigeria: Bills on hate speech and social media are dangerous attacks on freedom of expression - Amnesty International. See also Okegbile, J. (2023, July 18). Nigeria: Revisiting Nigeria’s Legal Framework On Hate Speech And Fake News Post 2023 General Elections. Mondaq. In recent years, Nigeria has grappled with the issue of hate speech and fake news, particularly in the context of elections. “The use of certain ‘foul’ and ‘hateful’ language and strategic misinformation can be highly divisive and can fuel violence, leading to significant harm to individuals, communities, and the country as a whole.” Amnesty International (2019, December 4). Nigeria: Bills on hate speech and social media are dangerous attacks on freedom of expression - Amnesty International. See also Okegbile, J. (2023, July 18). Nigeria: Revisiting Nigeria’s Legal Framework On Hate Speech And Fake News Post 2023 General Elections. Mondaq. In recent years, Nigeria has grappled with the issue of hate speech and fake news, particularly in the context of elections. “The use of certain ‘foul’ and ‘hateful’ language and strategic misinformation can be highly divisive and can fuel violence, leading to significant harm to individuals, communities, and the country as a whole.”
Constitutional Court, November 28, 2019, 2016Hun-Ma90 (Case on Restricting Online Media from Publishing Columns, etc. Written by Candidates for Public Official Election) (S. Kor.).
Constitutional Council, December 20, 2018, Decision no. 2018-773 DC (Fra.).