AI is changing the way America campaigns in the 2026 midterm elections. Political parties and organizations associated with them are creating such videos and advertisements with the help of artificial intelligence, in which opponents can be shown saying or doing things they did not actually do. To understand this shift, researchers at the Wesleyan Media Project analyzed election ads and used ad spending data from AdImpact.
The research revealed 220 election ads made with AI, costing $137 million, or about Rs. 1,200-1,300 crore was spent. According to the report, about 81% of the ads were from Republican candidates or groups affiliated with them, while the Democratic camp accounted for about 12%. This shows that AI is now being used on a large scale in election campaigns. Let us understand this trend in detail.
Fake photos and videos of opponents prepared with AI
The use of AI in America is no longer limited to scripting or creating illustrations for election ads. By copying the faces, voices and gestures of the leaders, such videos are prepared, which look realistic. This involves showing leaders saying things or doing things they did not actually do.
An example of this has come up in New York. Here, Republican gubernatorial candidate Bruce Blackman’s campaign produced a video. In which New York Mayor Johran Mamdani and Governor Cathy Hochul were shown walking and gardening together in a park. In the video, the two leaders were seen promoting a fictitious drug, dubbed the ‘Political Delusion Pill’.
Through the video, voters were taunted on their concerns about inflation and crime. It said that the side effects of this drug can include constant pocket burning and paying more taxes. This video may look real, but it is created with the help of AI. The ad also did not mention that the video is fake and created with artificial intelligence.
Over 80% of the ads are from Republican campaigns.
Researchers at the Wesleyan Media Project analyzed 220 ads to understand the use of AI in election campaigns. It was revealed that more than 80% of the ads were made by Republican candidates or political organizations affiliated with them. In comparison, the share of the Democratic camp was very small.
According to research, political parties are using AI to strengthen the image of their candidates and target their opponents. With the help of AI, it has become easy to show a leader with a different political face or to present such a picture of his views and political affiliations, which may differ from reality.
Another example of this was seen in Oklahoma. An ad supporting the Republican candidate there showed him hugging and holding hands with former Democratic President Hillary Clinton. Words like ‘Clinton Lover, Trump Hater’ were written in the video. It was intended to show the candidate as close to Clinton and an opponent of President Donald Trump.
This video was also created with AI, but there is no disclaimer providing this information. According to ad tracking company AdImpact, more than $200,000 was spent to air the ad on television in Oklahoma.
Mass promotion at low cost
AI has made the process of creating election ads easier and faster. Earlier, it cost time and money to shoot the video with the leader, engage the actors and shoot in different locations. Now with the help of AI tools such videos can be made in less time, in which the leader can be shown in any situation. This gives political parties an opportunity to advertise differently to different voter groups.
Promotional materials can be designed keeping in mind issues like inflation in some areas, crime in some areas and immigration in some areas. However, it can also be used to spread misinformation and tarnish the image of opponents. So the question now is not how cheap and effective AI is making election campaigning, but how it can help voters distinguish between genuine and fake content. These examples that emerged in the US midterm elections show that AI has become an important part of political campaigning. Transparency in election advertisements and identification of AI generated content may pose a major challenge to the democratic process in the future.
