Here is a 2,000-word news article in six paragraphs.
With South Africa’s local government elections now little more than a month away, the Electoral Commission of South Africa has taken an important legal step to protect the integrity of the vote. On 4 September, the commission gazetted the Electoral Code of Conduct Countering Disinformation in the Local Government Elections, commonly known as the Disinformation Code. The timing is significant: the country’s most fiercely contested local polls are scheduled for 4 November, and public confidence in ordinary citizens’ ability to separate fact from falsehood is slipping. According to the IEC’s latest Voter Participation Survey, the proportion of respondents who describe themselves as confident in recognising fake news fell from 62% in 2023 to 52% in 2025/26. Even more troubling, only 18% say they regularly verify election-related information before liking, sharing or reposting it. This erosion of confidence comes at a moment when disinformation is becoming harder to detect, largely because of generative artificial intelligence and increasingly hyper-local campaigning. The IEC’s new code is therefore not merely symbolic; it creates enforceable obligations for political parties and candidates and provides a formal mechanism for the public to report suspected disinformation. It is a recognition that the information environment around elections has changed fundamentally and that the old assumption that voters can simply tell truth from lies can no longer be taken for granted. At its core, the code is designed to place responsibility on those who produce and distribute political content, rather than leaving citizens to fend for themselves in a digital swamp. It covers everything from party manifestos and campaign advertisements to social media posts and AI-generated material. It applies to political parties, candidates and, importantly, the broader networks through which political content travels. The code obliges parties and candidates to verify information before sharing it, to label political advertising and synthetic content clearly, and to correct false or misleading claims within 36 hours. It also gives the IEC’s Real411 platform a central role as the official channel for complaints. In a country where disinformation has increasingly targeted specific wards and communities, the code is an attempt to create minimum standards of honesty in the public sphere. Whether it will be enough is another question, but its introduction marks an important moment in South Africa’s democratic evolution.
Although television remains the dominant source of voting information, the digital shift is unmistakable. The IEC’s survey found that 71% of respondents rely on television and 38% on radio for election-related information, while 23% now cite social media as a source, up from just 8% in 2021. That growth reflects a broader pattern: South Africans spend an extraordinary amount of time online. In 2025, the average daily online time was recorded at nine hours and 37 minutes, the highest in the world. Roughly three and a half of those hours are devoted to social media. The problem is that greater access to information has not translated into stronger verification practices. Only a small fraction of citizens regularly check what they read before passing it on. This gap between exposure and verification creates fertile ground for disinformation, especially in local elections where ward-level issues are often discussed in private WhatsApp groups and Facebook pages rather than on traditional platforms. The danger is not hypothetical. The World Economic Forum’s Global Risks Report 2026 ranks misinformation and disinformation among the top risks facing societies globally. In South Africa, the threat is becoming more localised and more difficult to monitor. Speaking at the Disinformation Dialogue 2026, IEC Chairperson Mosotho Moepya warned that the 2026 elections would likely see a shift from national misinformation and disinformation to “ward-specific deceptions” driven by generative artificial intelligence and hyper-local campaigning. His warning is consistent with research by the Institute for Security Studies, which has documented how disinformation and political messaging are increasingly targeted at specific communities and voter groups. The days when fake news was a broad, easily identifiable national narrative are giving way to a more insidious pattern: a false rumour here about a candidate, a manipulated image there about a development project, all circulating within closed groups where professional journalists and fact-checkers rarely venture. This hyper-localisation makes detection harder and makes the new code more necessary.
At the heart of this new landscape are community WhatsApp groups and Facebook pages, which have become central arenas for political discussion at ward and municipal levels. In these spaces, information travels quickly, context is stripped away, and verification is often limited. A screenshot, a voice note, or a short video can be forwarded to hundreds of people in minutes, and once it enters the group, it is frequently treated as truth. The research conducted by the Institute for Security Studies shows that political parties are well aware of this dynamic and are investing heavily in digital campaigning. Online political advertising and influencer-driven campaigns are becoming increasingly important features of election strategies. On one level, these tools can enhance political engagement by allowing parties to communicate directly with voters and by enabling communities to discuss issues that matter to them. But they also make it harder for voters to distinguish between authentic political discussion, sponsored content and coordinated influence campaigns. A post that appears to be a neighbour’s sincere opinion may in fact be a paid advertisement or the product of a sophisticated influence operation. The distinction between organic political conversation and engineered messaging is blurring. Artificial intelligence is compounding the problem. South Africa has not yet experienced a major election crisis linked to deepfakes, but AI-generated content is becoming cheaper, more accessible and harder to detect. Synthetic audio, in particular, is a growing concern: a fake voice note of a candidate saying something inflammatory could be created quickly and distributed widely before the victim has any opportunity to respond. The same technology that powers chatbots and image generators can be used to produce false but credible campaign material, and as the tools improve, the technical skill required to use them falls. This is why information integrity is no longer simply a media issue. It has become an integral part of citizenship, democracy and governance. Citizens need to know that the information they use to make voting decisions is reliable, and institutions need to be able to respond quickly when it is not. Technology, of course, is not inherently the problem. It can be used both to spread harmful falsehoods and to help voters access accurate information. South Africa’s challenge is to ensure that the tools designed to strengthen democratic participation evolve as quickly as the technologies used to manipulate it.
Much of the public conversation around disinformation focuses on social media users and their failure to verify content. But verification cannot be the sole responsibility of citizens. Professional and legacy media journalists remain one of democracy’s most important lines of defence against false and misleading information. They investigate claims, provide context and test competing narratives before they reach the public. In a local election, where a false allegation against a ward candidate can be devastating, the role of local journalists is especially important. Yet newsrooms are increasingly pressured to produce more content with fewer resources. Staff cuts, shrinking budgets and the demand for rapid digital output have changed the media environment. Still, there is evidence that standards are not necessarily being compromised. Newzroom Afrika, a South African news channel, told ISS Today that accuracy takes precedence over being first. The organisation applies a three-step verification process and labels fabricated AI-generated content as fake. These standard practices are reflected in the Press Council of South Africa’s guidance on the use of AI in journalism, which encourages newsrooms to be transparent about the tools they use and to verify the authenticity of user-generated content. The IEC, for its part, has expanded its efforts to strengthen information integrity and public trust. Voter education campaigns and public roadshows have been accompanied by journalist training, digital outreach and new tools enabling voters to verify information through official channels. These initiatives acknowledge that the media alone cannot solve the problem and that official institutions need to be proactive in providing reliable information. They also signal a broader shift in thinking: instead of simply responding to disinformation after it has spread, the focus is increasingly on preventing it from taking hold. The new code complements these efforts by creating a clear set of rules for political actors. It is a recognition that the information ecosystem is now too complex for citizens to navigate alone. Governments, media, political parties and technology companies all have responsibilities, and the burden must be shared.
Under the Disinformation Code, political parties and candidates are required to verify information before sharing it and to clearly label political advertising and AI-generated or synthetic content. They must also correct and retract false or misleading information within 36 hours. Anyone can report suspected election-related online disinformation through the IEC’s Real411 platform, which provides a structured route for complaints and follow-up. Crucially, the code extends responsibility for information integrity beyond official party statements to the wider networks through which political content circulates. This means that party supporters, community groups and even individual influencers can in principle be held to account for spreading disinformation, not just the official party machinery. This is an important innovation because much of the most damaging content in local elections is not produced by a party headquarters; it is shared and reshared by ordinary citizens who may not realise that what they are forwarding is false or manipulated. The code also places obligations on social media platforms, urging them to strengthen their responses to election-related disinformation. As political parties and candidates face greater expectations to take responsibility for the information they share, platforms should similarly improve their detection and removal of harmful content. The IEC chairperson has warned that breaches of the Disinformation Code carry the same legal consequences as violations of the core Electoral Code of Conduct. That is a significant statement. It means that the code is not merely a set of voluntary guidelines or a public relations exercise; it is a legally enforceable instrument that can be used to discipline parties and candidates who deliberately mislead voters. The prospect of legal consequences may encourage more caution in campaign communications, particularly in the final weeks before the vote. However, the effectiveness of the code will depend on enforcement. The IEC must be willing and able to act on complaints, and the public must be willing to use the reporting mechanism. There are also questions about the speed of enforcement in a fast-moving digital environment. A correction 36 hours after a false statement has spread may be too late to undo the damage. Nonetheless, the code represents a serious attempt to hold political actors accountable for the information they put into the world.
One of the code’s most obvious limitations is its duration. It remains in force only until the election results are declared, but the problem of disinformation can persist long afterwards. The allegations of vote rigging that followed the 2024 elections are a reminder that the period after voting is often when the information environment becomes most toxic. Losing parties and their supporters may use social media to cast doubt on the legitimacy of the results, and such narratives can take root even in the absence of evidence. South Africa will therefore need long-term information integrity solutions that extend beyond elections. This could include strengthening media literacy programmes in schools, investing in local journalism, developing clearer rules for political advertising online, and building the capacity of institutions to respond to future threats. The issue is not going away, and the next election will present its own challenges. In the immediate term, safeguarding local democracy is a shared responsibility, and citizens are an important first line of defence. Small, deliberate actions can strengthen democratic resilience. Challenging a suspicious post in a community WhatsApp group, helping a family member verify a candidate’s claim, or taking a moment to check the source of a viral message before sharing it are simple but meaningful acts. With six weeks left before voting starts, South Africans need access to information they trust. That requires a sustained commitment to verification from political parties, institutions, media organisations, platforms and citizens alike. The new Disinformation Code is an important step, but it is not a silver bullet. It is a tool, and like any tool, its value will depend on how consistently and effectively it is used. The real test will come on 4 November and in the days and weeks afterwards, when the health of South Africa’s local democracy will be judged not only by the conduct of the vote itself, but by the quality of the information environment in which that vote took place.
