EU AI law puts pressure on South African newsrooms

Transparency obligations under the EU AI Act have come into force in Europe, offering a reference point for South African newsrooms as AI becomes more embedded in journalism. The law regulates AI use in content creation, distribution and audience engagement, with requirements around labelling, transparency and accountability for harmful AI-generated content.

South African media already operate under professional standards overseen by bodies including the Independent Communications Authority of South Africa, the Broadcasting Complaints Commission and the Press Council of South Africa. Hendrik Sittig of the Konrad-Adenauer-Stiftung’s Media Programme for Sub-Saharan Africa said the EU framework does not stop newsrooms from using AI, but requires clearer disclosure and accountable editorial oversight.

A Centre for Information Integrity in Africa study found AI adoption in South African newsrooms remains uneven, often driven by individual interest rather than formal policy. Journalists raised concerns about secret use, unclear boundaries, plagiarism, weak handling of African languages and mistranslation of political or cultural terms.

The Press Council has urged editors to ensure AI-generated material is checked by human staff, fact-checked and disclosed to audiences, though its guidance is not binding. Industry figures argue that disclosure, explainability, training and audience trust must become central as newsrooms turn to AI while facing staff cuts and automation pressures.