In the report “AI as a weapon in election campaigns” in MIT Technology Review (issue 8/2024), Prof. Christoph Bieber, research professor at the Center for Advanced Internet Studies (CAIS), critically assesses the current debates on the influence of artificial intelligence on democratic elections. Against the backdrop of the 2024 super election year, the article warns against alarmist narratives that portray AI as an all-powerful tool for manipulating voters.
Prof. Bieber emphasizes that there is currently no reliable scientific evidence that AI-supported disinformation has demonstrably influenced election results. Although anecdotal examples from individual countries such as Czechia and India could be cited, there is a lack of systematic studies that could prove clear effects. The frequently used reference to allegedly clear research results therefore falls short. Rather, empirical research shows that political opinion-forming is a complex process in which technological means play only a minor role.
The article takes up Christoph Bieber’s assessment that although generative AI facilitates the production and dissemination of disinformation, this content is lost in a massive background noise of competing information. Structural factors such as trust in institutions, social polarization, and existing political attitudes continue to be decisive. An excessive focus on AI also carries the risk of distracting from deeper-rooted problems such as pressure on journalism, the erosion of democratic control mechanisms, or the strategic use of falsehoods by political actors.