South Africa 2024 General Election

An optimisation of the data-warehouse extract job was migrated to the live system at 06h15 on 31 May; at 06h30 the leaderboards at the results operations centres, the public website, the dashboard, the APIs and the feeds to media and political contestants all showed zeros, while internal management reports continued to display data. Displays were restored at about 08h45. The commission states that one line of test code left behind during the migration set the ready-to-display flag to zero for all results, that 2,614 voting district results for 1,213 voting districts were captured during the interrupted period, and that no results data was affected.

The Electoral Commission reports that from special voting on 27 May the voters' roll application on the devices randomly wrote duplicate records to local storage and transmitted them, in some cases more than 10,000 for a single transaction, congesting the application pools so that other devices responded slowly or not at all; the root cause was single-threaded execution on Android timing out during voter sessions. The African Union mission recorded malfunctioning devices, long queues and poor management of counting procedures; the African Centre for Governance mission recorded delayed openings, devices largely offline and some stations voting late into the night.

The Electoral Commission extended the section 55 objection deadline to 18h00 on 1 June 2024. It decided 110 objections, rejected 85, recorded three withdrawals and upheld 17, which produced 35 ordered recounts in the affected voting stations. It declared the result on 2 June 2024, four days after the close of voting and within the seven-day statutory limit. The African Union mission separately reports 579 objections and complaints received before the results were finalised, at least 10 authorised edit windows and 24 recounts during tallying.

Real411, the complaints platform run jointly by the Electoral Commission and Media Monitoring Africa since 2019, received 289 election-related complaints, against 99 in the 2019 election period and 50 in 2021. The commission's report states that the review panel concluded that 53% of the complaints fell within the elements of misinformation or disinformation, and that the rise in reports may reflect growing public awareness of the platform as well as a rise in the underlying content. Media Monitoring Africa's interim report of 31 May 2024, covering the roughly 200 complaints received to that point, put the proportion at 47% and found that X carried 63% of all complaints and 68% of the content found to be disinformation; X was invited to the commission's framework of cooperation with social media platforms and did not join, and the commission records it as maintaining no policies or standards on misinformation and disinformation; the commission names Meta, Google, Microsoft and TikTok among the technology companies it worked with. One of the four content categories recorded was material intended to undermine the credibility of the Electoral Commission and its commissioners; the commission's own report describes cyberattacks, social media harassment, attacks on officials, unlawful access to storage facilities and intimidation of election officers.

On 20 May 2024 the Constitutional Court held unanimously in CCT 97/24 that his contempt sentence of more than 12 months engaged section 47(1)(e) of the Constitution, that the provision turns on the sentence imposed rather than the time served after presidential remission, and that he is disqualified until five years have elapsed since completion of the sentence. The judgment overturned the Electoral Court's decision of 26 April 2024, which had gone the other way. The disqualification ran to the candidate, not the party: uMkhonto weSizwe contested the election and won 58 seats.

The party asked the Electoral Court to declare the result not free and fair and invalid and to order a fresh election within 90 days (case 34/2024EC); it withdrew that application and costs were awarded to the Electoral Commission. A second application on the same grounds (39/2024EC) was still before the Electoral Court as of May 2025, and at a hearing on 17 June 2026 the party argued a narrowed case confined to the results-display interruption; judgment was reserved and no decision had been reported when this page was reviewed on 29 August 2026. The African Transformation Movement's application (35/2024EC) was dismissed without an oral hearing on the ground that it produced no credible evidence and that challenging the result before exhausting the section 55 procedure was premature; the African Restoration Alliance's appeal against the rejection of its objections (32/2024EC) was dismissed without an oral hearing for unjustified delay and little prospect of success. No court has upheld an allegation that the count was falsified.

The Electoral Commission's published account is that the functional part of the results system remained available throughout, that capturing at municipal sites continued, and that only the display of the data was disturbed. uMkhonto weSizwe argues that optimising the live system during counting was itself unjustified interference. The Constitutional Court dismissed the party's June 2024 application to interdict the first sitting of the National Assembly (CCT 178/24) on the grounds that the challenge did not engage its exclusive jurisdiction and that the delay in launching it was not justified, rather than on the merits of the allegation.

The African Union, SADC and African Centre for Governance missions and the Electoral Commission's own report identify domestic disinformation, cyberattacks on the commission and platform governance gaps as the information-integrity problems of this election, not foreign operations. A study published in December 2024 that specifically examined foreign information manipulation and interference alongside generative AI found most false content originating from ordinary domestic sources. Absence of a documented finding is not the same as a finding of absence: no forensic attribution exercise covering this election has been published.

The December 2024 study for the German Council on Foreign Relations found a limited presence of AI-driven disinformation, with most false content coming from traditional sources rather than AI tools, and no notable decline in public trust in media attributable to AI-generated content beyond isolated deepfake images and impersonations. Media Monitoring Africa's platform data points the same way: the reported harms were overwhelmingly ordinary false claims about processes, parties, results and the Electoral Commission. Neither source is an authoritative refutation in the sense of a court finding, audit or forensic review, so this is recorded as the direction the available research points rather than as a settled matter.