As South Africa prepares for the 2026 Local Government Elections, the Centre for Human Rights at the University of Pretoria, in partnership with UNESCO and the Social Media 4 Peace (SM4P) Coalition, convened a Pre-Election Information Integrity Dialogue to examine the growing challenges posed by online misinformation, disinformation, xenophobic narratives and artificial intelligence (AI) to democratic participation. The dialogue brought together representatives from civil society, the media, regulatory institutions, academia and the technology sector to explore practical responses to information disorder and its implications for electoral integrity, freedom of expression and public trust.
The dialogue opened with remarks by Dr Sabelo Gumedze, who highlighted the importance of safeguarding democratic processes as South Africa approaches the 2026 Local Government Elections. Soren Larsen, representing the European Union, underscored the importance of initiatives that promote information integrity and democratic participation. Dr Phathiswa Magopeni, Chair of the Social Media 4 Peace (SM4P) Coalition and Vimbai Mataruse of UNESCO also contributed to the opening of the dialogue, highlighting the importance of collaboration in addressing online harms while protecting freedom of expression and democratic participation.
Examining the risks of online mobilisation and anti-migrant narratives
The dialogue opened with a presentation of research findings by Kyle Janse from Centre for Analytics and Behavioural Change (CABC), examining online anti-migrant mobilisation and the spread of potentially harmful content across X, TikTok, Facebook and YouTube between 23 June and 4 July 2026. The research focused on xenophobic hate speech, incitement to violence and misinformation targeting foreign nationals in the context of online mobilisation around 30 June. It examined not only the content circulating online but also how platforms responded when potentially harmful material was formally reported. The findings highlighted the risks associated with online mobilisation in a country with a history of xenophobic violence and unrest. Although the 30 June mobilisation did not result in comparable nationwide violence, the research cautioned against interpreting the absence of large-scale violence as evidence that the underlying risks had disappeared. Of approximately 113 posts formally escalated to platforms, action was reported on around one-third. TikTok recorded the strongest response, while the research presentation highlighted significant gaps in reported action on other platforms. The findings raised questions about the effectiveness of existing reporting mechanisms, the responsiveness of platforms during periods of heightened risk and the need for dedicated escalation channels during crises.
Recommendations included establishing crisis-response protocols, creating accessible escalation mechanisms for trusted civil society organisations, explicitly addressing migrant status in platform policies and strengthening transparency around content moderation decisions. The research also emphasised the importance of distinguishing harmful content from legitimate criticism of immigration policies, political commentary, protest and satire. Such distinctions are essential to ensuring that efforts to address online harms do not unnecessarily restrict freedom of expression.
Campaigning, migration and the integrity of electoral information
The first panel examined the relationship between political campaigning, migration and information disorder. Moderated by Dr Ivy Gikonyo, the discussion featured Dr Phathiswa Magopeni and Kyle Janse. Panellists explored how misleading narratives can spread through WhatsApp groups and other social media platforms, particularly when they exploit public concerns about housing, water, electricity and other essential services. Such narratives can move rapidly from private messaging groups into wider public debate, making it difficult to establish their origins and assess their potential impact. The discussion highlighted the growing sophistication of AI-generated content, including manipulated images and videos, which can make it increasingly difficult for audiences to distinguish authentic material from fabricated content. Participants stressed that journalists and members of the public must go beyond relying on visual clues and undertake more rigorous verification of information and its sources.
The role of local and community media was another key theme. Panellists emphasised the importance of reporting that reflects local realities, uses accessible language and enables communities to understand electoral processes and assess claims that directly affect their lives. The discussion also examined the responsibilities of political parties, media organisations and digital platforms in addressing inflammatory campaign narratives. Participants raised concerns about the normalisation of anti-migrant rhetoric and the risks of political actors exploiting public frustration to mobilise support. Importantly, the panel highlighted the need to interrogate the context in which migration-related narratives emerge, including how media coverage can amplify claims without adequately examining their evidence or underlying assumptions.
The broader message was that protecting electoral integrity requires more than correcting false claims. It also requires responsible political communication, contextualised reporting and meaningful scrutiny of narratives that may deepen social divisions.
AI, platform accountability and electoral information integrity
The second panel focused on AI and electoral information integrity, with Yolakazi Mfuto moderating a discussion featuring William Bird of Moxii Africa, Reggy Moalusi of the South African National Editors’ Forum (SANEF), Advocate Pansy Tlakula, Chairperson of the Information Regulator, and Prince Tshweu, Founder of the Local Government Elections World Cup (#LGEWC). Advocate Tlakula joined the discussion from the United States.
The panel explored the implications of AI-generated and manipulated content, algorithmic recommendation systems, platform transparency and the need to strengthen public AI literacy as the elections approach. Participants considered how digital platforms can amplify misleading political narratives and create information environments in which users are repeatedly exposed to content that reinforces their existing views. These dynamics can affect public perceptions of political parties, electoral institutions and the credibility of electoral processes.
The discussion also highlighted concerns about online attacks against political figures, particularly women, as well as xenophobic and anti-foreigner narratives entering electoral discourse. The use of AI to manipulate political content added another layer of complexity to the information integrity challenge. Moalusi highlighted SANEF’s work to strengthen journalistic capacity ahead of the elections, including training more than 1,000 journalists across all nine provinces. The training addresses information integrity, ethical reporting and journalists’ safety, reinforcing the importance of building resilience beyond election periods. Bird also shared insights from election-related monitoring through Real411, drawing attention to reported complaints involving disinformation, hate speech, incitement to violence and attacks on electoral institutions. The discussion underscored the importance of scrutinising unsupported allegations against electoral bodies, which can undermine public confidence in democratic processes.
Prince Tshweu introduced the Local Government Elections World Cup (#LGEWC), a non-partisan, no-budget civic-engagement initiative designed to encourage public interest in the 2026 Local Government Elections. The initiative uses a 16-team knockout tournament format, with participating political parties seeded according to their share of the 2024 national vote. Members of the public decide each match through polls hosted on a WhatsApp Channel, while short videos on TikTok, Instagram and YouTube generate interest in the match-ups. Each match is accompanied by a sourced information dossier covering the Constitution, the parties’ track records, manifestos, 2024 vote shares and available public polling. Where information is not publicly available, this is explicitly indicated. The initiative aims to use an accessible and engaging format to encourage people to learn more about political parties, assess their performance and research their choices before voting. Its final was scheduled for 30 October 2026, ahead of the elections on 4 November.
Advocate Tlakula’s contribution brought a regulatory perspective to political campaigning and personal information. She explained that political parties and independent candidates may process voters’ personal information for legitimate campaigning purposes, but such activities remain subject to the Protection of Personal Information Act (POPIA). Lawful processing requires compliance with applicable conditions, including requirements relating to the lawful collection of information, purpose limitation, data minimisation, accuracy and appropriate safeguards.
These considerations become particularly important in digital campaigning environments, where personal information and AI-driven systems can influence how political messages are targeted and delivered. The discussion highlighted the need to promote responsible campaigning, strengthen accountability and equip voters to engage critically with information circulating online.
Moving from dialogue to practical action
As finally summarised by Dr. Tendai Mbanje, across both panels, participants emphasised that electoral information integrity cannot be secured by any single institution. Effective responses require sustained cooperation between government, electoral bodies, regulators, media organisations, civil society, technology companies and academic institutions. The discussions also highlighted the importance of balancing interventions against harmful content with the protection of freedom of expression. Addressing misinformation and incitement requires context-sensitive approaches that distinguish legitimate political debate from content that threatens democratic participation and human dignity. As the 2026 Local Government Elections approach, the dialogue reinforced the need to move beyond identifying risks towards implementing practical, coordinated responses. These include strengthening platform accountability, improving reporting and escalation mechanisms, investing in media literacy, supporting local journalism and ensuring that political campaigning respects fundamental rights.
The Centre for Human Rights and its partners will continue to engage stakeholders through the Social Media 4 Peace Coalition to advance information integrity, protect freedom of expression and strengthen democratic participation in South Africa’s evolving digital environment.
Watch Recording
For more information, please contact:
