Restrictions on freedom of expression

The Constitution of the Republic of Serbia protects freedom of expression as a fundamental right, but it also constitutes a framework for its restriction. In that sense, freedom of expression could be limited by others’ rights and reputation and the obligation to uphold the authority and objectivity of the court and protect public health, the morals of democratic society, and the national security of the Republic of Serbia (Art. 46(2)).

This general provision, in fact, has the same meaning as Art. 10 of the European Convention of Human Rights. In that sense, the general provision restricting the freedom of expression has two mutual functions. One is the state’s duty to enable any citizen to express their opinion without fear of being punished for it or being put in a different position compared to the unlike-minded. It also includes the accessibility of someone’s expression. The other one is the state’s duty as well, but it refers to the duty to protect the rights of others.

In order to fulfil both aims, the provisions of several laws prescribe certain content that could result in the violation or jeopardization of others’ rights. These are not numerus clausus cases. In that sense, if the content is not explicitly rec-ognized as inappropriate, this does not mean it is allowed. In fact, the court that will decide on eventual disputes determines whether the expressed content violates someone’s rights or which one of the confronted rights requires protection in a spe-cific situation.

According to a survey on rights violations on the Internet,⁶ the most frequent violations are in relation to threatening content, compromising security, insults, and

6 The research was conducted by the Share Foundation through continuous monitoring of violations in online communication. Results in different categories of monitoring are available at https://mon-itoring.labs.rs/.

unfounded accusations.⁷ The affected parties are mostly citizens. However, the re-verse is also true, as citizens comprise the majority of attackers. Other analyses consider regional rights violations in the digital environment. The results show that hate speech and discrimination are involved in almost half of the reports of Internet-based digital violations.⁸

Various actions and content could represent the violation and jeopardization of human rights, national security, and the morals of democratic society, but at the same time, this content represents the line between freedom of expression and ex-pression that is punishable by law.

For the purpose of understanding the restriction of the freedom of expression, violations that are prevalent among unlawful actions committed on the Internet will be explained.

2.3.1. Threatening content and compromised security

Content that contains threats to attack another person’s life or body could en-danger that person’s safety. In order for Internet content to be considered en-dangerous, it must cause fear in the person to whom it is directed, but it must also be directly stated and achievable. In each case, a competent body (usually a court) must termine whether these conditions are met. It is not always easy to correctly de-termine whether something is an expression of thoughts that are appropriate to jeop-ardize someone’s right (because of a lack of seriousness). Moreover, it is sometimes not obvious which expression among similar ones has the potential to endanger someone’s safety, which one causes fear, and finally, whether the attacked person is too sensible for the content. All relevant issues should be evaluated according to the circumstances of the particular case.

7 There are seven different categories of digital right and freedom breaches and several sub-categories under each that have been under monitoring: 1) information security breaches, including making con-tent unavailable through technical means, destruction, and the theft of data and programs, computer fraud, unauthorized access – unauthorized alterations and insertions of content, disabling control over an online account or content; 2) information privacy and personal data breaches (publishing infor-mation about private life, illegal interception of electronic communications, breaches of citizens’ per-sonal data,  illegal perper-sonal data processing, breaches of information privacy in the workplace, other breaches of information privacy); 3) pressures because of expression and activities on the Internet (publishing falsehoods and unverified information with the intention of damaging another’s reputa-tion, insults and unfounded accusations, threatening content and endangering of security, hate speech and discrimination, freedom of expression on the Internet and the workplace, pressures because of publishing information); 4) manipulation and propaganda in the digital environment (creating fake accounts and paid promotion of false content, content manipulation and organized reporting on social media; changes to or removal of content that is in the public interest; placement of commercial content as news, other manipulation in the digital environment); 5) holding intermediaries liable (pressures because of user-generated content); 6) blocking and filtering of content (blocking/filtering on the network level, algorithmic blocking or suspension of content); 7) other breaches of digital rights and freedoms not included in the categories previously defined. Available at: https://monitoring.labs.rs/.

8 Data available at: https://monitoring.bird.tools/.

To illustrate, the Supreme Court of Serbia, in a decision during a criminal pro-cedure, stated that the defendant, as one of the members of a group advocating violence against individuals belonging to the gay population or organizing a public gathering, sent a threat, the meaning of which was the deprivation of life or at least bodily injury of the participants in an event. The threat was sent through an electronic medium intended for the general public. That means that the threat was available to everyone, that is, a wide range of people, including the mentioned persons to whom the threat was addressed in a certain place. Bearing this in mind, the threat, given the specified circumstances under which it was addressed and could have been known, could objectively create anxiety, insecurity, and fear regarding the personal safety of the members of that circle of persons, who, as passive subjects and possible victims, are indicated in the indictment, in a manner that sufficiently identifies them as members of the target group to which the threat is addressed.⁹

Notwithstanding the outcome of the above case, the court indicates which ele-ments of any statement made on the Internet, on a social network to be precise, are relevant to making the decision.

2.3.2. Insults and unfounded accusations

Expressions on the Internet that violate others’ reputation and honor shall be considered unlawful. However, expressions with a negative connotation are not them-selves enough to jeopardize someone’s right to dignity, reputation, and honor. De-valuing the attacked person is an important element of violation. For that reason, the person toward whom the insult is directed has to be determined. In addition, the statement ought to be insulting according to objective criteria (such as the behavior of the society or group, common opinion on the quality of the statement, and the moral values of the particular group or society), not just the insulted person’s perception.¹⁰

Insults occurring in the media or in the public forum are a qualified crime, and hence stricter punishment is handed down.¹¹

2.3.3. Hate speech and discrimination

The Republic of Serbia’s Constitution and laws prohibit discrimination. Moreover, it seems that much attention is paid to the violation of others’ rights, especially the right of equity. On one side, the reason for this treatment is the fact that equity is a

9 Judgement of the Supreme Court of Serbia, Kzz 59/11, dated 31 August 2011.

10 In Lepojić vs. Serbia, and Filipović vs. Serbia, the European Court of Human Rights stated that author-ities have a duty to suffer stronger criticisms and insults, and they need to be much more tolerant than ordinary citizens. This statement was contrary to the opinion of Serbian courts, which claimed that the honor of the authorities is more important than the same personal right of others. Kršikapa, 2008, p. 3.

11 Art. 170, par. 2 Criminal Code of the Republic of Serbia, Official Gazette of the Republic of Serbia, No. 85/05, 72/09, 111/09, 121/12, 104/13, 108/14, 94/16, 35/19. More on this: Petrašinović, 2020, pp. 469–481.

precondition of the exercise of other human rights.¹² There is no value in protecting freedom of religion, for example, if religions are not treated equally by the com-petent law or in practice. The other reason for this legally developed restriction of freedom of expression is the divergence of behavior or content that could result in discrimination.

According to the Law on Prohibition of Discrimination of the Republic of Serbia,¹³ discrimination is any unjustified differentiation, inequitable conduct, or misconduct in relation to persons or groups or members of their families or persons close to them, directly or indirectly, if this behavior is based on race, color, ancestry, citi-zenship, nationality or ethnic origin, language, religious or political beliefs, gender, gender identity, sexual orientation, property status, birth, genetic characteristics, health status, disability, marital or family status, conviction, age, appearance, mem-bership in political, trade union, and other organizations, and other real or assumed personal characteristics.¹⁴

As could be concluded from the definition provided in the mentioned provisions, discrimination can be conducted in any way and can be based on any personal char-acteristic. The condition that must be met in order to consider an active or passive behavior inequitable is unjustified differentiation among citizens. In the context of this work, the terms ‘justified’ and ‘unjustified’ will not be explained. Attention will be paid to the specific manifestation of discrimination through the exercise of freedom of expression. It is related to hate speech.

With the development of the Internet and social networks, opportunities for spreading intolerant, discriminatory, and hateful content have become greater. Bearing this in mind, legal and institutional mechanisms for grappling with communicative aggression on the Internet have been developed during the last de-cades and within the current century in general.¹⁵ The Law on Public Information and Media from 2014 and the Code of Journalists of Serbia define the media’s ob-ligation to edit all content contained in their publications, including readers’ com-ments below the texts on online portals. The media are obliged to beware of and prevent the spread of hate speech and aggressive communication through its publica-tions. Social networks, forums, and other online platforms, however, are not subject to these rules, and it is more difficult to edit discriminatory content and take mea-sures against creators of hate speech. 

The fact that hate speech is regulated by several laws illustrates Serbian society’s sensibility with regard to the consequences of destructive communication. In that sense, breach of the prohibition of hate speech, as a specific kind of discrimination, is treated differently in comparison to other human rights breaches.

12 Etinski, 2013, pp. 51–67; Etinski, 2005, pp. 325–352.

13 Official Gazette of the Republic of Serbia, No. 22/2009.

14 Mršević, 2014, pp. 13–16.

15 Intensive anti-discriminatory action has been expected after conflicts in the Balkans region. For more on the development of the anti-discriminatory approach: Saša Gajin (ed.), 2010; Zubčević et al., 2017; Reljanović, Matić, and Ilić, 2010; Saša Gajin (ed.), 2015; Rašević, 2014.

Specifically, apart from the criminal offense that includes hate speech as a sub-stantial element of the crime, the protection the general anti-discriminatory law provides should be implemented in the dispute via the court decision.

Specific rules are applied in discrimination disputes. Fault for breach (intention or negligence) is not required for the determination of hate speech. The only claim in which fault is relevant is damage compensation. In that case, the absence of the intention to spread or encourage hate, aggression, or violence can release someone from the liability to compensate for damages. This is not the case in other claims.¹⁶

Further, regarding civil protection, the essential difference between the pro-cedure for protection against discrimination and hate speech according to the Law on Prohibition of Discrimination and the procedure according to the Law on Public Information and Media¹⁷ is reflected in passive legitimacy, i.e., the person against whom the request is directed. Whereas according to the Law on Prohibition of Dis-crimination, the protection of the endangered or violated right to equality is imple-mented via a demand directed against the discriminator, in lawsuits under the Law on Public Information and Media, the demand is directed against the editor-in-chief of the media publication.¹⁸ The editor-in-chief’s liability arises from their responsi-bility for the content they edit. This is regardless of who authored the information, whether they are a journalist or someone else. In other words,  the editor-in-chief’s responsibility exists not only when the author of the information is a journalist, but also when the hate speech published in the media comes from others. In online media publications, indiscriminate publication of readers’ comments on a certain topic in which the ban on hate speech is violated is not rare.¹⁹

Protection under criminal law is narrower because the criminal offense exists in cases of inciting national, racial, and religious hatred and intolerance.²⁰ The of-fender’s guilt depends on proving that their intent in this case was aimed at inciting religious hatred and intolerance among nations. To be precise, the fact that there was a verbal or physical attack on another people, nationality, ethnic community, or religion does not mean that the subjective (and substantial) element of this crime has been demonstrated.

16 According to Art. 43 of the Law on Prohibition of Discrimination, through a lawsuit, the plaintiff may demand: 1) imposing a ban on an activity that poses the threat of discrimination, a ban on proceeding with a discriminatory activity, or a ban on repeating a discriminatory activity; 2) that the court should establish that the defendant has treated the plaintiff or another party in a discrim-inatory manner; 3) taking steps to redress the consequences of discrimdiscrim-inatory treatment; 4) com-pensation for material and non-material damage; 5) that the decision passed on any of the lawsuits referred to in previous items be published.

17 Law on Public Information and Media, Official Gazette of Republic of Serbia, No. 83/2014, 58/2015, 12/2016.

18 Art. 103 of the Law of Public Information and Media.

19 Rašević, 2018, pp. 1309, 1310.

20 Art. 128 of the Criminal Code.

2.3.4. Breach of the right to privacy

Disclosure of personal data, details about private life, photos, and other infor-mation without the concerned person’s consent violates the right of privacy.²¹ This right is a focus of modern legislature, precisely because of the increased incidence of uncontrolled use of gathered personal information. In most cases, it is about abuse.

The increasing popularity of the Internet, especially social networks, has led to a more intensive consideration of privacy protection.  Specifically, social network websites contain user information such as age, relationship status, income, and in-formation about close family members, as well as registered users’ addresses. The most common abuses and violations of privacy using data available on social net-works are identity theft and the manipulation of personal data, which are related to employment and the misuse of photos on the Internet.²² A large number of social network services store personal data about users, so that users do not have to re-enter them when they want to use them later, e.g., for online shopping, booking over Internet sites, or simply revisiting the same website.²³

Simultaneously with the development of artificial intelligence and machine learning, expectations regarding the appropriate improvement of the legal instru-ments pertaining to personal data protection have arisen. Ethical issues arising from the application of artificial intelligence are a special aspect of the human rights debate amidst the so-called Fourth Industrial Revolution.  This seems to be even more dominant if the fact that humans spend the majority of their daily lives under a certain type of (in)direct monitoring, intentionally or spontaneously, is considered.

Smart devices connected to the Internet, the network of public and private surveil-lance cameras, and automated decision-making based on online behavior history are just some of the currently recognized sources of dispute between citizens, public policy, and industry.

The misuse of photos on the Internet is a form of privacy violation that occurs when photos from social network users’ accounts are used and displayed without their consent. A photo of any person can be displayed in a way that may harm them on a personal level. Over time, social network users could download the photo, re-sulting in a large number of users having an opportunity to see the photo, share it, or even worse, make a photomontage that could be further misused. In addition, some social networks (e.g., Facebook) note in their terms of use that they reserve the right to publish user information or share it with other companies, lawyers, courts, government agencies, etc., if deemed necessary.²⁴

21 Judgement of the Court of Appeal in Belgrade, Gž3 77/18, as of 29 March 2018.

22 Identity theft consists of the unauthorized use of personal data (date of birth, current residence, telephone number, occupation, friends, personal pictures), which have become publicly available.

Identity theft on the Internet is a form of fraud committed by computer users involving gathering personal and financial information through a fake e-mail or website. Đukić, 2017, pp. 99–116.

23 Midorović and Sekulić, 2019, pp. 1158–1159.

24 Diligenski and Prlja, 2018, pp. 27–31.

In this context, children are a particularly vulnerable category, and for this reason, wide, decisive action on the part of competent bodies is more than welcome.