Scenario 19: Hate speech

State A uses a social media platform headquartered in State B to incite racial and religious hatred against an ethnic and religious minority in its own territory. The resulting violence escalates into a non-international armed conflict involving cyber operations between State A and members of the ethnic and religious minority who organise themselves into an armed opposition group. The scenario analyses whether the incidents amounted to violations of international law, including international human rights law, international humanitarian law and international criminal law.

Keywords
International human rights law, international humanitarian law, prohibition of genocide, non-international armed conflict, attacks against persons, individual criminal responsibility, hate speech, incitement of violence

Facts
[F1] Security agents of State A use a globally operating social media platform to spread hate speech against an ethnic and religious minority in the same State. The social media platform does not have a physical office in the territory of State A, but it directs its operations from its global headquarters in State B. Over a period of several years and with increasing intensity, the security agents use fake accounts, news pages, and trolls on the social media platform to spread false, inflammatory information about the religious and ethnic minority, labelling it as “ethnically inferior”, “terrorist”, “criminal” and “existential threat to State A that must be eliminated”. A high-ranking security official also posts on his profile that “the problem regarding the ethnic and religious minority will be solved soon”.

[F2] An independent international fact-finding commission establishes that this spread of inflammatory misinformation was a decisive factor in contributing to mass violence targeting the ethnic and religious minority, including killings, inhumane treatment, torture, and rape. It finds that these acts were perpetrated by non-State organisations representing the ethnic and religious majority and often aided or abetted by the military forces of State A. The government of State A does not take any measures to stop the spread of the misinformation by its security agents. Despite calls by civil society for the social media platform and State B to take action, the social media platform reacts only slowly in taking down the relevant pages and fake accounts. State B emphasises that the imposition of laws against hate speech would violate the freedom of speech as guaranteed under its constitution.

[F3] In response to the mass violence, the members of the ethnic and religious minority form an armed opposition group, Group C, which is headquartered in and assumes control over a province in State A that is predominantly populated by the ethnic and religious minority. Group C, led by several commanders appointed by the group’s General Staff, launches armed counterattacks against both the military of State A and associated non-State organisations with the proclaimed aim “to liberate the ethnic and religious minority from the systematic and violent oppression by State A”. As a result, those different actors engage in protracted armed violence, which causes a large-scale displacement of persons. To stop the spread of inflammatory misinformation about the ethnic and religious minority, hackers who are part of the armed opposition group attack government servers in State A, which are used by the security agents to disseminate the misinformation, by infiltrating them with ransomware. Because of the attack, the government servers lose their functionality, which also affects civilian infrastructure such as hospitals.

[F4] State A and State B are parties to the International Covenant on Civil and Political Rights (ICCPR), the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD), the Convention on the Prevention and Punishment of the Crime of Genocide (the ‘Genocide Convention’), the 1949 Geneva Conventions and Additional Protocols I (AP I) and II (AP II) thereto, as well as the Rome Statute of the International Criminal Court (the ‘Rome Statute’).

Examples

 * Hate speech in India (since 2017)
 * Hate speech in Myanmar (since early 2010s)
 * Syria’s ‘social media war’ (since 2011)

Legal analysis
For a general overview of the structure of analysis in this section, see Note on the structure of articles.

[L1] Focusing on the spread of hate speech on social media, the analysis first examines whether State A and State B could be held internationally responsible for violations of international human rights law and the prohibition of genocide. In view of the possible escalation of the violence into an armed conflict, the analysis then discusses the international responsibility of State A for violations of international humanitarian law. Lastly, the individual criminal responsibility of State A’s security agents and Group C is considered.

[L2] The scenario assumes that the conduct of the security agents was attributable to State A. Pursuant to Article 4 of the ILC Articles on State Responsibility for Internationally Wrongful Acts, the conduct of any State organ, irrespective of its position within the State, its functions and its character as an organ within the central government or territorial unit, shall be considered an act of that State. Security agents of a State undoubtedly qualify as organs of that State. Therefore, any act of State A’s security agents is attributable to State A. According to Article 7 of the ILC Articles on State Responsibility, this is so even if the security agents exceed their authority or contravene instructions.

Violations of international human rights law applicable to cyber operations
[L3] Although the security agents of State A used a social media platform that is headquartered in State B, the spread of inflammatory information took place in the territory of State A. The posts by State A’s security agents distinguished the minority inter alia based on ethnic origin. In addition, as the independent international fact-finding commission found, the dissemination of incendiary misinformation on the social media platform contributed to mass violence targeting the minority, which impeded their enjoyment of the right to life, the freedom from torture and possibly other human rights. The spread of inflammatory misinformation could thus be characterized as “racial discrimination” as defined in Article 1 of the ICERD.

[L4] In terms of the specific obligations under the ICERD, it is conceivable that State A’s spread of inflammatory misinformation about the ethnic and religious minority violated its obligation set out in Article 2(a) of the ICERD. The provision prescribes that “[e]ach State Party undertakes to engage in no act or practice of racial discrimination against persons, groups of persons or institutions and to ensure that all public authorities and public institutions, national and local, shall act in conformity with this obligation”. By posting incendiary content and by aiding or abetting the discriminatory acts of non-State organisations representing the ethnic and religious majority, State A could have also violated its obligation “not to sponsor, defend or support racial discrimination by any persons or organizations” under Article 2(b) of the ICERD.

[L5] Moreover, State A might be internationally responsible for its failure to act. Pursuant to Article 2(c) of the ICERD, State A was under an obligation “to review governmental, national and local policies, laws and regulations and to amend, rescind or nullify any laws and regulations which have the effect of creating or perpetuating racial discrimination”. Such laws and regulations would include any internal policies against the ethnic and religious minority pursued by the State’s security apparatus. As State A failed to take measures to stop the spread of discriminatory information via the social media platform, it could have also violated its obligation under Article 4 of the ICERD to “condemn all propaganda and all organizations which are based on ideas or theories of superiority of one race or group of persons of one colour or ethnic origin” and to “undertake to adopt immediate and positive measures designed to eradicate all incitement to, or acts of, such discrimination”. Moreover, this conduct could also imply a breach of Article 20(2) of the ICCPR, which states that ‘any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law’, read together with State A’s obligation under Article 2(1) of the ICCPR to ‘respect and to ensure to all individuals within its territory and subject to its jurisdiction the rights recognized in the [Covenant], without distinction’.

[L6] Obligations under international human rights treaties do not directly bind corporations such as the social media platform. To the extent that the social media platform did not take sufficient measures to prevent and end the spread of hate speech about the ethnic and religious minority, the platform could only be said to have acted inconsistently with its ‘responsibility’ to respect human rights under the non-binding UN Guiding Principles on Business and Human Rights.

[L7] It could also be examined whether State B is internationally responsible for having failed to prevent and stop the spread of inflammatory misinformation. State B’s international responsibility might be based be on a breach on Article 2(d) of the ICERD, which prescribes that “[e]ach State Party shall prohibit and bring to an end, by all appropriate means, including legislation as required by circumstances, racial discrimination by any persons, group or organization”, and on the above-mentioned Article 4 of the ICERD (see para. L5). Moreover, Article 20(2) of the ICCPR includes the broader prescription that “[a]ny advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law”. State A disseminated inflammatory misinformation about the ethnic and religious minority over the course of several years. During that time, State B could have taken and enforced legislative measures to prevent or halt the incitement to racial or religious hatred through a social media platform based on its territory. According to the interpretative guidance provided by the respective United Nations committees, the prescriptions under Article 4 of the ICERD and Article 20 of the ICCPR are fully compatible with the right of freedom of expression. This suggests that it should be possible to enact hate speech laws that can be reconciled with the freedom of speech under the constitution of State B.

[L8] Nonetheless, it is controversial whether the ethnic and religious minority in State A was subject to the jurisdiction of State B, which is necessary to trigger the applicability of the pertinent human rights obligations. The issue of the extraterritorial application of human rights obligations is governed by the relevant human rights treaties. There is “no restriction of a general nature in [I]CERD relating to its territorial application”. It could thus be argued that the obligations of State B under ICERD are triggered by the failure to prevent or stop the spread of inflammatory misinformation. In contrast, the ICCPR will likely not be applicable in the present scenario. State B only had spatial jurisdiction over the social media platform, which was operating from its territory. The social media platform might have played a detrimental role in interfering with the ethnic and religious minority’s right to non-discrimination, which underlies the obligation to enact legislation against discrimination in Article 20 of the ICCPR. However, the minority was not subject to the physical control of State B.

Violations of the prohibition of genocide through cyber means
[L9] The spread of inflammatory misinformation via the social media platform might be in violation of State A’s obligations under the Genocide Convention and the prohibition of genocide under customary international law. The killings, inhumane treatment, torture, and rape committed against the ethnic and religious minority could potentially qualify as genocide under Article II(a)–(c) of the Genocide Convention.

[L10] While the acts in question were committed by members of non-State organisations representing the ethnic and religious majority, State A could be responsible for direct and public incitement to commit genocide under Article III(c) of the Genocide Convention. Evidence of an intent to destroy the ethnic and religious minority in whole or in part may be found in the labelling of the ethnic and religious minority on the social media platform as “ethnically inferior”, “terrorist” and “criminal” and an “existential threat to State A that must be eliminated” by the security agents of State A, as well as in the post by a high-ranking security official that “the problem regarding the ethnic and religious minority will be solved soon”. The relevant posts were public because they were made on the social media platform. The directness criterion may be fulfilled even if the State in question does not explicitly call for genocide. However, State A would have to “specifically provoke another to engage in a criminal act”, which is more than “a mere vague or indirect suggestion”, i.e. there would have to be a causal connection. The post that the ethnic and religious minority was an “existential threat to State A that must be eliminated”, within the broader context of the campaign of spreading ethnic and religious hatred over a period of several years, could potentially fulfil this criterion. As an inchoate crime, direct and public incitement to genocide does not necessitate that genocide actually follows.

[L11] State A could have also breached its obligation to prevent genocide or the prohibition to be complicit in its commission. While complicity in genocide presupposes actual knowledge of and aid or assistance in an imminent or ongoing genocide, the obligation to prevent genocide only requires the State to know of the serious risk that genocide may be committed. As established by the independent international fact-finding commission, the spread of the inflammatory misinformation decisively contributed to potentially genocidal acts committed against the ethnic and religious minority. That finding indicates that State A had sufficient knowledge of the risk in question to suggest that it may have violated its duty to prevent genocide. In addition, not only did the military forces of State A fail to stop these acts from happening, but they also aided and abetted the perpetrators. State A may thus also have violated the prohibition to be complicit in the commission of genocide. Unlike incitement to genocide, a violation of the obligation to prevent genocide and complicity in genocide require that genocide actually took place.

[L12] Other States parties to the Genocide Convention might equally be under an obligation to prevent the potential genocide in State A. As highlighted by the calls by civil society, State B arguably had the capacity to stop the spread of inflammatory misinformation, for example, by ordering the social media platform to take down the illegal content. Given the status of the prohibition of genocide as a jus cogens norm, all States and international organizations are under an obligation cooperate to bring to an end through lawful means any serious breach of that norm, and not recognize as lawful a situation created by the serious breach, nor render aid or assistance in maintaining that situation. Accordingly, State B’s failure to act may have amounted to a violation of the said obligations under international law.

Violations of international humanitarian law applicable to cyber operations
[L13] The hostilities between State A and Group C will likely meet the definition of a NIAC involving cyber operations, which triggers the applicability of international humanitarian law. The hostilities between State A and Group C led to protracted armed violence that caused a large-scale displacement of persons, thus reaching the level of intensity required for a NIAC. In addition, the attacks by Group C on the government servers significantly damaged civilian infrastructure, which further speaks to the severity of the fighting. Group C also had the necessary organisation to fulfil the second criterion for a NIAC. The group was headquartered in a particular province of State A, which it largely controlled. It spoke with one voice, and it acted under a unified command with a hierarchical structure led by a General Staff that clearly proclaimed the strategic aim of the group, namely “to liberate the ethnic and religious minority from the systematic and violent oppression by State A”.

[L14] It is unlikely that State A would have violated the prohibition to attack the civilian population through its involvement in the spread of hate speech against the ethnic and religious minority. Incitement to racial and religious hatred by cyber means per se does not qualify as an attack against the civilian population. Such incitement might rather be compared to other non-violent cyber operations such as psychological cyber operations and cyber espionage, which are not considered to be attacks. State A could also submit that the spread of misinformation was a permitted ruse of war, for example, by leading Group C to falsely assume that that military operations are imminent or on-going.

[L15] Even if the dissemination of inflammatory posts on the social media platform during the NIAC did not qualify as an attack, the threat of a kinetic attack by cyber means could violate the prohibition to spread terror among the civilian population. A credible proclamation by a party to the conflict that “threaten[s] the annihilation of civilian populations” qualifies as a prohibited threat of violence. In that regard, it is material that the security agents of State A called the ethnic and religious minority “criminal”, “terrorist”, “ethnically inferior” and, in particular, an “existential threat to State A that must be eliminated”. These proclamations, insofar as they were made in the context of an armed conflict, may therefore have violated the prohibition of threats of violence aimed at terrorizing the civilian population.

[L16] In addition, State A could have breached Common Article 1 of the Geneva Conventions, which provides that “[t]he High Contracting Parties undertake to respect and to ensure respect for the present Convention in all circumstances”. This obligation implies that they “may neither encourage, nor aid or assist in violations of the Conventions”, including provisions that are applicable in NIACs such as the prohibition of acts of violence to life and person under Common Article 3. The spread of misinformation by State A with a view to inciting violence against the ethnic and religious minority during the NIAC could be understood as such encouragement to violate the Geneva Conventions.

Violations of international criminal law applicable to cyber operations
[L17] The security agents and the military forces of State A as well as the members of Group C could incur individual criminal responsibility for their acts committed by cyber means.

[L18] By spreading inflammatory misinformation on the social media network, the security agents of State A might be individually criminally responsible for directly and publicly inciting genocide pursuant to Articles 6 and 25(3)(e) of the Rome Statute. Among the different “channels of communication” to incite genocide, social media could be considered particularly effective. The labelling of the ethnic and religious minority as an “existential threat to State A that must be eliminated” and the post that “the problem regarding the ethnic and religious minority would be solved soon” might support a finding that the security agents, including their commanders, had the required intent to provoke or prompt to commit genocide. While the killings, rape and torture of the ethnic and religious minority could be considered to amount to genocide on the different grounds of Article 6 of the Rome Statute, the commission of genocide is not required for a conviction for incitement to genocide.

[L19] Considering the mass violence against the ethnic and religious minority in State A, it could be examined whether the security agents and members of the military forces of State A are responsible for crimes against humanity pursuant to Article 7 of the Rome Statute. In this context, the incitement to racial and religious hatred on the social media platform could arguably be part of a widespread and systemic attack directed against a particular part of civilian population, which included different acts that qualify as crimes against humanity listed in Article 7(1) of the Rome Statute. Those acts happened over a period of several years and seemed to have been part of a State or organisation policy to commit an attack against the civilian population, notably the ethnic and religious minority.

[L20] An argument could be made that the spread of hate speech, and in particular the incitement to violence, amounted to persecution under Article 7(1)(h) of the Rome Statute. The posts on the social media platform took place “in connection with any act referred to in this paragraph 1 or any crime within the jurisdiction of the Court”, namely killings, torture and inhumane treatment and rape committed against the ethnic and religious minority. To constitute persecution, the incendiary social media campaign against that minority, which clearly is an identifiable group or collectivity, should also meet the required threshold of gravity as defined in Article 7(2)(g) of the Rome Statute. It would have to be determined to which extent the posts labelling the ethnic and religious minority as “existential threat to State A that must be eliminated” deprived the minority of its fundamental rights, such as the right to security. In this regard, the independent international fact-finding commission’s conclusion that such posts were a decisive factor in contributing to mass violence targeting the ethnic and religious minority is relevant. It then remains to be established whether State A’s security agents had discriminatory intent and were aware of the detrimental impact of those posts.

[L21] It might also be argued that State A’s security agents, by posting of inflammatory misinformation, could be criminally responsible for instigating crimes against humanity as a mode of liability prescribed by Article 25(3)(b) of the Rome Statute. Instigation means “prompting another to commit an offence”, which implies a causal nexus between the instigation and the commission of the crime. In the present scenario, the relevant crimes under Article 7(1) of the Rome Statute, which were committed by the non-State organizations, include murder (subparagraph (a)), torture (subparagraph (b)) and rape (subparagraph (g)). State A’s security agents did not order the crimes, but Article 25(3)(b) also encompasses solicitation and inducement, which could possibly cover the inflammatory posts on the social media platform. While State A’s military forces might have “aided” in the commission of such crimes against humanity, the continuous dissemination of inflammatory information by the security agents could also be characterized as “abetting” in terms of Article 25(3)(c) of the Rome Statute.

[L22] For acts committed during the NIAC, the security agents and the military forces of State A could potentially incur individual criminal responsibility for war crimes. In particular, by posting false inflammatory information about the ethnic and religious minority on the social media platform, the security agents could be responsible for instigating non-State organisations representing the ethnic and religious majority to commit acts of violence to life and person under Article 8(2)(c)(i) and committing rape under Article 8(2)(e)(vi) of the Rome Statute. Moreover, the military forces might be responsible for abetting the commission of such crimes.

[L23] Commanders and other superiors in State A might be criminally responsible for not preventing the acts that constitute or contribute to genocide, crimes against humanity and war crimes. It is likely that the commanders and superiors in State A knew or, owing to the circumstances at the time, should have known that their forces were committing or about to commit the relevant crimes. Moreover, they failed to take all necessary and reasonable measures within their power to prevent or repress the commission of the crimes, i.e. to stop the posting of inflammatory misinformation and the aiding and abetting of mass killings, torture and inhumane treatment and rape.

[L24] It may further be investigated whether members of Group C could have committed war crimes by attacking protected objects (i.e. the hospitals in State A) under Article 8(2)(e)(iv) of the Rome Statute, provided they did not qualify as military objectives. Nonetheless, the target of the ransomware attacks was not the hospitals, which could mean that the hackers did not have the required intent to attack the protected object. Accordingly, the members of Group C would likely not incur individual criminal responsibility under this ground.

Checklist

 * International human rights law
 * Did State A violate the prohibition of incitement to hatred and the prohibition of racial discrimination?
 * Is State B under a positive obligation to prevent the spread of hate speech extraterritorially pursuant to the applicable international human rights treaties?
 * Prohibition of genocide
 * Is State A internationally responsible for direct and public incitement to genocide?
 * Did State A violate its obligation to prevent genocide and/or was it complicit in genocide?
 * Is State B (and possibly other members of the international community) under an obligation to prevent genocide?
 * International humanitarian law
 * Did the violence between State A and Group C qualify as a NIAC?
 * Did the spread of hate speech by cyber means violate the prohibition to attack civilians in an armed conflict?
 * Did the spread of hate speech violate the prohibition of threats of violence aimed at terrorizing the civilian population during armed conflict?
 * Did State A encourage violations of the Geneva Conventions by spreading hate speech on social media?
 * International criminal law
 * Did the potential perpetrators fulfil the material elements (actus reus) and mental elements (mens rea) of the offence?
 * Did the potential perpetrator commit the offence individually, jointly or through another?
 * Did commanders or superiors incur criminal responsibility (a) for ordering cyber operations that constitute international crimes, or (b) for failing to take all reasonable and available measures to prevent the commission of those crimes or to punish those responsible if they knew or should have known that their subordinates were committing the crimes?

Bibliography and further reading

 * Sina Alavi (ed), The Council of Advisers’ Report on the Application of the Rome Statute of the International Criminal Court to Cyberwarfare (August 2021).
 * Kai Ambos, ‘International Criminal Responsibility in Cyberspace’ in Nicholas Tsagourias and Russell Buchan (eds), Research Handbook on International Law and Cyberspace (Edward Elgar 2015).
 * James Banks, ‘European Regulation of Cross-Border Hate Speech in Cyberspace: The Limits of Legislation’ (2011) 19 Eur J Crime CrL & CrJ 1.
 * M. Cherif Bassiouni, Introduction to International Criminal Law (2nd edn, Martinus Nijhoff 2014).
 * William H Boothby, The Law of Targeting (OUP 2012).
 * Antonio Cassese and Paola Gaeta, Cassese’s International Criminal Law (3rd edn, OUP 2013).
 * Yoram Dinstein, Non-International Armed Conflicts in International Law (CUP 2014).
 * Cordula Droege, ‘Get Off My Cloud: Cyber Warfare, International Humanitarian Law, and the Protection of Civilians’(2012) 94 IRRC 533.
 * Stefanie Farrior, ‘Molding The Matrix: The Historical and Theoretical Foundations of International Law Concerning Hate Speech’ (1996) 14 BerkeleyJIntlL 3.
 * Andrey Fino, ‘Defining Hate Speech – A Seemingly Elusive Task’ (2020) 18 JICL 31.
 * Shannon Fyfe, ‘Tracking Hate Speech Acts as Incitement to Genocide in International Criminal Law’ (2017) 30 LJIL 523.
 * Paola Gaeta (ed), The UN Genocide Convention: A Commentary (OUP 2009).
 * Gregory S Gordon, ‘Formulating a New Atrocity Speech Offense: Incitement to Commit War Crimes’ (2012) 43 LoyUChiLJ 281.
 * Gregory S Gordon, ‘Hate Speech and Persecution: A Contextual Approach’ (2013) 46 VandJTransnatlL 303.
 * ICRC, International Humanitarian Law and the Challenges of Contemporary Armed Conflicts: Recommitting to Protection in Armed Conflict on the 70th Anniversary of the Geneva Conventions (ICRC 2019).
 * Mark Klamberg (ed), Commentary on the Law of the International Criminal Court (Torkel Opsahl Academic EPublisher 2017).
 * Marco Roscini, ‘Gravity in the Statute of the International Criminal Court and Cyber Conduct that Constitutes, Instigates or Facilitates International Crimes’ (2019) 30 CrimLF 247.
 * Michael N Schmitt (ed), Tallinn Manual 2.0 on the International Law Applicable to Cyber Operations (CUP 2017).
 * Wibke K Timmermann, ‘The Relationship between Hate Propaganda and Incitement to Genocide: A New Trend in International Law Towards Criminalization of Hate Propaganda?’ (2005) 18 LJIL 257.
 * Wibke Timmermann and William A. Schabas, ‘Incitement to Genocide’ in Paul Behrens and Ralph Henham (eds), Elements of Genocide (Routledge 2013).
 * Richard A Wilson, ‘Inciting Genocide with Words’ (2015) 36 MichJIntlL 277.

Contributions

 * Scenario by: Christiane Rexilius
 * Analysis by: Christiane Rexilius
 * Reviewed by: Marjolein Busstra, Emma Irving, Tilman Rodenhäuser