Global Media Rights and Responsibilities
Senior 6 · General and Religious Studies
Global Media Rights and Responsibilities
Lesson 2 of 3
25 min
On 3 December 2003 the International Criminal Tribunal for Rwanda (ICTR) gave judgement in what became known as the Media Case. Two of the accused had founded the radio station RTLM; the third edited the newspaper Kangura. Before and during the 1994 Genocide against the Tutsi, these outlets spread hatred against Tutsi, and RTLM broadcast calls to violence, including names of people to be hunted. The tribunal convicted all three of direct and public incitement to commit genocide, among other crimes. On appeal in 2007 some convictions were overturned and the sentences reduced, but the incitement convictions linked to RTLM and Kangura stood.
Reading its decision, the Trial Chamber said: "The power of the media to create and destroy fundamental human values comes with great responsibility. Those who control such media are accountable for its consequences."
For Rwandans this is not theory. It is the reason media responsibility is taught in every school, and the background to this lesson's questions: who is responsible for what we read, hear and watch, and when may speech be limited?
Definition
Global media: media whose content crosses borders, such as international broadcasters, streaming services and social-media platforms.
Journalism ethics: the principles of good conduct for people who gather and publish news, such as accuracy, fairness, independence and care to avoid harm.
Misinformation: false information shared by someone who believes it is true. Disinformation: false information created or shared on purpose to deceive.
Hate speech: expression that attacks or demeans people because of who they are, such as their origin, religion, sex or disability. Incitement is hate speech that urges others to discriminate, be hostile or use violence.
Censorship: the suppression or control of what people may say, publish, read or watch, by a government, a company or another authority. Self-censorship: holding back one's own speech from fear of the consequences.
Content moderation: a platform's review and removal of posts that break the law or its own rules.
Algorithm: a set of rules a computer follows. On platforms, algorithms choose which posts each user sees first.
Deepfake: an image, audio or video made or altered with artificial intelligence (AI) so that it seems to show a real person saying or doing something they never did.
Codes of journalism ethics around the world share four core duties. The table uses the headings of the widely used Code of Ethics of the Society of Professional Journalists (2014).
| Duty | What it requires | Example |
|---|---|---|
| Seek truth and report it | Verify facts, name sources where possible, separate news from opinion | Confirming a road-accident death toll with the police before broadcasting it |
| Minimise harm | Weigh the public's need to know against the harm to people; take special care with children and victims | Not showing the face of a child witness |
| Act independently | Refuse gifts and pressure that could bias reporting; declare conflicts of interest | A sports reporter not covering a club that pays her |
| Be accountable and transparent | Admit and correct errors quickly; explain choices to the audience | Publishing a correction as visibly as the error |
In Rwanda, Law N° 02 of 2013 regulating media gives a journalist three obligations: to inform; to educate the population and promote leisure; and to defend the freedom of information and analyse and comment on it (Article 5). The daily conduct of journalists is regulated by a Media Self-Regulatory Body, while the national utilities regulator also oversees audio, audio-visual media and the internet (Article 4). Media for children carry extra restrictions (Article 7).
A newspaper editor decides what millions read. Today a platform's algorithm makes similar choices for billions of people, every second, often by showing whatever keeps people scrolling. Anger and fear keep people scrolling. So responsibility now extends beyond journalists (Figure 1).
Case study: Myanmar, 2017–2018. During the violence against Rohingya Muslims in 2017, posts spreading hatred circulated widely on Facebook, which for many people in Myanmar was the internet. The United Nations Independent International Fact-Finding Mission on Myanmar reported in 2018 that Facebook had been a useful instrument for those seeking to spread hate. The company later accepted that it had not done enough, hired reviewers who read Burmese and removed accounts of military leaders. The case shows the duties in Figure 1: a platform operating in a country must understand its languages and act fast when speech turns into incitement.
Artificial intelligence. Generative AI can write news-like text and make realistic images and voices in seconds. It also makes confident errors, and it can be used to make deepfakes. The UNESCO Recommendation on the Ethics of Artificial Intelligence (2021), adopted by all 193 UNESCO member states, asks for transparency, human oversight, fairness and protection of privacy. The European Union's AI Act (2024) requires deepfakes to be disclosed as artificially generated (Article 50). Rwanda adopted a National Artificial Intelligence Policy in 2023, and its Law N° 58 of 2021 relating to the protection of personal data and privacy limits how anyone, including AI developers, may collect and use personal data.
| Form | What happens | Possible justification | Danger |
|---|---|---|---|
| Prior censorship | Material must be approved before it is published | Rare: for example, military secrets in wartime | Officials decide in advance what citizens may know |
| Punishment after publication | Courts punish unlawful speech, such as incitement or defamation | Protects people's lives, safety and reputations | Vague laws can punish honest criticism |
| Platform moderation | A company removes posts or accounts | Stops incitement, fraud and abuse at scale | Private firms decide without public rules or appeal |
| Internet shutdowns | Networks or apps are switched off | Governments cite violence or exam cheating | Everyone loses news, payments and emergency calls |
| Self-censorship | People keep silent from fear | None: it is a symptom, not a policy | Problems go unreported and unsolved |
International law answers the question with a test. Under Article 19, paragraph 3, of the International Covenant on Civil and Political Rights, a restriction on expression must be (1) provided by law, (2) for a legitimate aim, such as the rights of others, national security, public order, or public health or morals, and (3) necessary and proportionate, that is, the least restrictive way to reach that aim. Article 20 goes further: advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence "shall be prohibited by law". Rwanda's Constitution follows the same logic in Article 38, and its Law N° 60 of 2018 on prevention and punishment of cyber crimes makes it an offence to knowingly publish rumours online that may incite fear, insurrection or violence (Article 39).
How do you tell offensive speech, which a free society tolerates, from incitement, which it must stop? The United Nations Rabat Plan of Action (2012) gives six factors: the context; the speaker and his or her influence; the intent; the content and form; the extent of its reach; and the likelihood that it will cause harm, including how soon.
Worked example 1: applying the three-part test
During tense local elections in a country, a popular musician with two million followers posts a video telling fans that a named religious group "must be driven out of our town before Sunday". The platform removes it within an hour. Was this justified?
Provided by law? Incitement to hostility or violence is prohibited by law, as Article 20 of the Covenant requires, and the platform's published rules ban it.
Legitimate aim? Yes: protecting the lives and rights of the group's members and public order.
Necessary and proportionate? Apply the Rabat factors: a tense context, an influential speaker, a clear call to act against a named group, a huge audience and a deadline that makes harm likely and soon. Removing one post is the least restrictive effective step; closing the whole platform would not be.
Conclusion. The removal was justified. A post that only criticised the group's beliefs, without any call to harm, would not meet the threshold.
Worked example 2: a deepfake two days before a vote
An audio clip spreads in which a candidate seems to admit to bribery. Experts find it was made with AI. Who should do what?
Every person with a phone is now a publisher. The SIFT method, devised by the digital-literacy researcher Mike Caulfield, gives four quick moves.
| Move | What you do |
|---|---|
| Stop | Notice your emotion. Strong anger or excitement is a signal to slow down. |
| Investigate the source | Who published this? Search for the source itself, not only what it says about itself. |
| Find better coverage | Are trusted outlets reporting the same claim? If none are, be wary. |
| Trace to the original | Find the first version of the quote, photo or video. Is it cut, old or from another country? |
Tip
Three-part test in one line: "lawful, legitimate, necessary". Use the three words as the paragraphs of any essay on whether a limit on speech is justified, then add the Rabat factors when the speech is alleged to be incitement.
Warning
Do not call all limits "censorship" or all speech "freedom". Strong answers distinguish. Stopping incitement to genocide is a duty under international law; jailing a journalist for an accurate report is a violation. The test, not the label, decides.
Important
Discuss. Should global platforms be held legally responsible for every post their users publish, as a newspaper is for its articles? Consider the scale (billions of posts a day), the risk that platforms will over-remove lawful speech to stay safe, and what happened in Myanmar.
Crash Course: "Check Yourself with Lateral Reading", Navigating Digital Information #3, presented by John Green (about 14 minutes). Watch for how professional fact-checkers open new tabs to check a source instead of reading it top to bottom.
Practice
Sources and credits
Written by Ganzaa: Figure 1, the tables, the worked examples and the exercises.
