American election campaigns are entering a new phase with the increasing use of artificial intelligence tools in the political advertising industry and the direction of election messages. The competition is no longer limited to television advertisements and digital publications, but has extended to automatically generated videos and images, and attempts to influence the information provided by chatbots to voters.
Hot Topics
This shift raises questions about the limits of legitimate political use of technology, the risks of misinformation, and the ability of electoral laws and institutions to keep pace with rapidly evolving tools.
One of the most notable changes in this year’s US midterm elections is the introduction of AI-generated content into political advertising, whether through producing scenes that did not happen, fabricating statements on the tongues of candidates, or using real images and characters in fabricated contexts.
Is artificial intelligence redrawing the job map in the technology sector?
Examples documented by US media reveal the use of these technologies in state and congressional election campaigns. In Texas, Republicans ran an ad featuring Democratic Senate candidate James Talarico making statements he hadn't actually recorded, after artificial intelligence superimposed his image and voice. Ads targeting Democratic Senator Jon Ossoff in Georgia also appeared, using fabricated scenes mocking him.
In New York, election ads used AI-generated images of Governor Kathy Hookall and Mayor Zahran Mamdani in situations that did not occur, including depicting them working together in a satirical political scene.
Data from the Wesleyan Project, which monitors political ads, indicates that more than 164 AI-generated or enhanced political ads were monitored during the election season, with spending exceeding $80 million. More than 80% of these ads were linked to Republican candidates or their supporters. These figures pertain to the ads monitored by the project and do not represent the total election spending on artificial intelligence.
The problem is that some of these advertisements do not necessarily try to convince the public that the scene is real, but rather use exaggerated or satirical images and situations to attract attention and establish a negative impression of the targeted candidate. This presents a broader concept than traditional deepfake, as the content can be clearly artificial, but it is still able to influence political debate and the circulation of stereotypes.
Tags:
technology