Global Experts Warn AI Misinformation Threatens Election Integrity

Government agencies and research organizations in the United States and South Africa are warning that artificial intelligence bias and misinformation could compromise the integrity of upcoming elections. In Bakersfield, California, panelists at the Real Election Talk: AI and Your Vote event cautioned voters against relying on chatbots for cognitive thinking, noting that these tools often reinforce user biases by being overly agreeable.

To counter these risks, experts recommended the SIFT method—stopping, investigating sources, finding better coverage, and tracing claims—and encouraged the use of FactCheck.org. These efforts to educate voters on identifying fabricated content are mirrored by institutional initiatives in other regions.

In South Africa, the Council for Scientific and Industrial Research partnered with the Government Communication and Information System and other stakeholders to host a virtual media roundtable on October 28, 2026. The discussion focused on how AI-generated audio, video, and images make it difficult to distinguish authentic content from fabrications. Participants, including Africa Check and Media Monitoring Africa, discussed using data science and inter-organizational collaboration to identify emerging narratives and safeguard public information.