- Emojis can be weaponised to insult, threaten, sexualise, exclude or conceal harmful communication.
- Context determines meaning: an emoji that appears harmless can contribute to cyberbullying, harassment or grooming depending on how it is used.
- South African law focuses on harmful conduct, rather than making specific emojis illegal, with legislation covering cybercrime, harassment, hate speech and child protection.
- Children need stronger digital literacy, including an understanding of consent, privacy, manipulation, online harassment and how coded communication can cause harm.
- Platforms, parents, educators and regulators share responsibility for protecting children as digital communication evolves beyond easily identifiable words and phrases.
Skull, clown, ‘laughing face’ and peach emojis may appear innocent at first glance, yet each can carry meanings that reach far beyond the symbols themselves.
Among young people, emojis and other forms of coded communication (memes, GIFs, etc.) have become a constantly evolving modern daily digital language. Usually, they signal humour, irony, intimacy or belonging.
But what happens when emojis are weaponised to insult, threaten, sexualise or deliberately exclude others? In these cases, the potential for emojis to function as a form of online harm becomes increasingly difficult to dismiss.
An emoji is not inherently harmful. Its meaning is socially produced and shaped by context, relationships and cultural norms. Yet this ambiguity can also enable harmful behaviour to be disguised or communicated in ways that are not immediately apparent to others.
For example, a degrading comment may be framed as a joke, a sexual reference can be expressed through seemingly innocuous symbols and bullying can be conveyed through a combination of images that appear harmless to outsiders but carry a clear message for the intended recipient.
Beyond words and symbols
This highlights that online harm extends beyond individual words or symbols. It may involve context, repetition, power dynamics, and impact. Also, a laughing emoji attached to a humiliating photograph, for instance, may intensify cyberbullying, while repeated coded sexual messages directed at a child may contribute to grooming or harassment.
In these contexts, the emoji functions not as the source of harm itself, but as a vehicle through which harmful communication can be expressed, normalised and even concealed.
Evidence from South Africa suggests that these concerns cannot simply be dismissed as hypothetical. UNICEF’s Disrupting Harm in South Africa study (2022) surveyed 1 639 internet-using children aged 9 to 17. In the tracking, 34% said someone had made sexual comments about them that made them uncomfortable; 24% had been sent sexual images they did not want; and 7% said someone had shared sexual images of them without their consent.
The study highlights the vulnerability of children to different forms of technology-facilitated sexual exploitation and abuse. The graph above does not show that emojis cause sexual harassment or exploitation.
Rather, it illustrates the wider communicative environment in which coded language operates. Harmful communication does not always announce itself through explicit language.
In her scholarship, Sonia Livingstone, Professor of Social Psychology at the London School of Economics and Political Science (LSE), highlights how digital and everyday life have become increasingly intertwined, noting that “it is becoming hard to draw the line between offline and online”.
Online and offline
Her observations are particularly relevant to children whose social lives move between school, home, WhatsApp groups, gaming communities and social media. An online humiliation can become an offline school conflict, and an offline relationship can become the channel through which digital harassment occurs. Sometimes, it is hidden in plain sight, unseen by parents, teachers and automated moderation systems.
South African law does not make an emoji, or a particular combination of emojis, illegal. Instead, it regulates certain forms of conduct and harm communicated electronically. The Cybercrimes Act 19 of 2020, for example, criminalises particular malicious communications.
Section 14 deals with electronic messages intended to incite damage to property or violence, Section 15 addresses electronic messages that threaten people with such harm and Section 16 criminalises the unlawful and intentional disclosure of an intimate image without the person’s consent.
The Children’s Bill (B70D-2003) defines ‘abuse’ broadly to include any form of harm or ill-treatment deliberately inflicted on a child, expressly encompassing bullying by another child and exposing or subjecting a child to behaviour that may cause psychological or emotional harm.
Section 28 of the Constitution, as reflected in the Constitutional Court’s discussion of Children’s Rights, protects children from maltreatment, neglect, abuse and degradation and requires that the child’s best interests be paramount in every matter concerning the child.
Harmful media
The Films and Publications Act 65 of 1996 protects children by regulating the classification, distribution and exhibition of potentially harmful content. It criminalises the creation, possession and distribution of child sexual abuse material and requires internet service providers to implement measures to restrict access to prohibited content.
The Film and Publication Board (FPB) enforces these provisions to safeguard children from harmful media.
South Africa’s Protection from Harassment Act 17 of 2011 provides another relevant legal framework. The Act recognises electronic communication that causes harm, or makes a complainant feel that they may be harmed, as potentially constituting harassment.
It allows a complainant to apply for a protection order, and the electronic communication can be considered as part of the harmful conduct.
The Prevention and Combating of Hate Crimes and Hate Speech Bill (B9D–2018) defines ‘communication’ in Section 1 to include displays, written or visual material, representations, references and electronic communications, thereby possibly encompassing forms of digital expression.
More importantly, Section 4(1) provides that a person commits hate speech where they intentionally communicate material in a manner that could reasonably be construed as demonstrating a clear intention to be harmful or to incite harm and to promote or propagate hatred based on one or more prescribed grounds.
Emojis not regulated
Real411, a public complaints platform run by Moxii Africa, provides a practical mechanism for addressing digital ‘offences’ such as hate speech, incitement, misinformation, disinformation and harassment. Its Digital Complaints Committee assesses complaints under its Code of Conduct and within South Africa’s constitutional and legal framework.
Crucially, its approach does not depend on identifying specific ‘bad words’. Harassment may instead be established through persistent or serious unwanted conduct that demeans, humiliates or creates a hostile or intimidating environment.
At the same time, Real411’s framework balances protection from harm with constitutional protections for Freedom of Expression, meaning offensive, unpopular, satirical, or ambiguous expression is not automatically unlawful. This is especially important because the boundary between humour, commentary and harmful conduct can be contested.
The existing Children’s Rights protections clearly do not regulate emojis. What is important to note, however, is that the law is emphatically concerned with the underlying harms that digital media can facilitate.
Section 16 of the Cybercrimes Act, the Films and Publications Act, and the Prevention and Combating of Hate Crimes and Hate Speech Bill do not depend entirely on whether harmful electronic communication uses conventional words.
Contextual meaning
Legal significance depends on contextual meaning. An emoji’s potential to cause harm depends not only on the emoji itself, but also on how it is used and understood within a particular context. A threat expressed through text, an image or a string of emojis can raise legal questions depending on its content, the intention of the communicator and the circumstances.
Real411 illustrates a broader principle that digital harm cannot always be understood by examining individual content in isolation. Thus, an emoji that appears harmless in isolation can, as part of a broader pattern of messages, images, or symbols, contribute to hate speech, harassment, or cyberbullying.
Livingstone challenges the assumption that growing up with technology automatically makes children digitally competent. She argues that the concept of the “digital native” can obscure significant differences in children’s skills, experiences, and circumstances. In her view, digital fluency should not be equated with digital literacy.
Familiarity with trending emojis, for example, does not necessarily indicate an understanding of consent, manipulation, privacy, harassment, or the potential consequences of sharing content.
Excessive surveillance
For parents and educators, this creates a difficult challenge as they are expected to monitor a form of communication that they may not understand. Excessive surveillance can undermine children’s privacy and trust, while doing nothing may leave serious harm unseen. Digital literacy education therefore needs to go beyond screen-time rules.
Young people need to understand how context changes meaning, how humour can become humiliation, and how forwarding or participating in harmful content can have consequences. But responsibility cannot rest with families alone. Platforms themselves have become central to the debate.
Australia introduced a minimum age requirement of 16 for social media access in December 2025, requiring platforms to implement effective age-verification measures to prevent underage users from accessing their services.
In contrast, South Africa’s Minister of Communications and Digital Technologies, Solly Malatsi, recently stated that the South African government would not introduce an outright ban on social media access for children under 16.
Meta’s recent settlement with attorneys general from 48 US states, Washington, D.C., and some U.S. territories to strengthen protections for teenagers on Facebook and Instagram demonstrates the growing pressure on technology companies to take greater responsibility for children’s online experiences.
Parental controls
Meta agreed to pay up to US$18 billion and introduce measures including stronger age assurance, a two-hour daily limit for users under 18, a midnight-to-6 am access block, stronger parental controls and limits on some social-comparison features. The safeguards are intended to apply by default and will reportedly operate in the U.S. for most of the next decade.
Yet the settlement also exposes the limitations of a safety model focused on time limits and parental controls. Some critics argue that such measures do not fully address recommendation algorithms and other design features that may contribute to harm.
The debate is therefore shifting from “how much time are children spending online?” to the more difficult question of “what kind of environment are they entering when they go online?”
That distinction matters greatly for coded communication. An automated system may identify an explicit threat relatively easily. But emojis are much harder to interpret. They can be used affectionately in one conversation and in a threatening or sexual manner in another.
Digital environment
A moderation system that removes every ambiguous emoji would be impractical, while one that overlooks it could leave victims unprotected. In South Africa, linguistic diversity, the digital divide, and varying levels of parental supervision shape young people’s online experiences.
Children who communicate multilingually, using a mix of vernacular with emojis, may use forms of expression that automated systems struggle to interpret accurately.
The answer, then, is not to declare emojis dangerous. It is to recognise that digital harm evolves faster than the vocabulary used to describe it. Emojis can facilitate, intensify and conceal harmful behaviour, but they are rarely the whole story.
The real issue is the digital environment in which they acquire meaning, an environment where children, parents, educators, regulators, governments and technology companies share responsibility for safety.
South Africa’s challenge, therefore, is to develop a more nuanced form of digital literacy that considers not only what an emoji means, but also who is using it, in what context, towards whom, and with what effect. And perhaps the law needs to keep asking the same question.
Because if harmful conduct can hide behind codes, symbols and seemingly harmless digital language, then protecting children cannot depend solely on whether an adult or an algorithm recognises the meaning behind an emoji.
Banathi Mgqoboka is an award-winning, Cape Town-based multi-platform journalist, writer and researcher working at the intersection of contemporary social realism and critical media analysis, bridging academic and non-academic critical discourse. He holds a Master’s Degree (MA) from Rhodes University’s School of Journalism and Media Studies, where he was a research scholar in the National Research Foundation (NRF) ‘Licence to Talk’ Project.
