Political deepfakes are already circulating well before the 2026 midterm elections, raising fresh concerns about how generative AI could shape voter perceptions. What began with fake photographs has evolved into manipulated video and, more recently, synthetic audio, giving bad actors new ways to imitate the voices and appearances of public figures.
Scammers have long attempted to extract money and personal information from unsuspecting people, but generative AI is lowering the barrier. With a single click, they can now produce convincing content that looks and sounds like a real person. This practice, known as AI impersonation, is a growing problem. Celebrities, public figures and politicians all rank among the most frequently targeted individuals.
AI-generated media has become commonplace, driven by tools from major technology companies including OpenAI, Google and Meta. Image and video generators can create and edit material with ease, requiring no advanced editing skills. That accessibility allows fake content to be produced and distributed at scale.
Understanding the Scale of the Threat
Bobby Ford, chief strategy and experience officer at Doppel, framed the challenge as the three V’s: volume, velocity and vectors. “Before, it would take a threat actor a certain amount of time, as well as resources, in order to create that deepfake,” Ford said. “Take that same amount of time, and it’s done at much faster clips, so the volume (and velocity) increases.” He added that new vectors such as video, audio and email have expanded as AI tools become increasingly multimodal.
The purpose of synthetic media is not always to produce a flawless imitation. Frequently, it is designed to make people believe fabricated content is genuine, manipulating emotions and overriding careful judgement. That capacity to influence beliefs carries particular weight in politics.
AI Is Already Embedded in Politics
During the 2024 primary presidential election, thousands of voters in New Hampshire received robocalls appearing to come from Joe Biden, urging them not to vote the night before the contest. The calls were an AI deepfake, and many voters could not tell the difference. Although the Biden campaign attempted to warn people, the damage had already been done.
Political campaigns “may not have the tools for cybersecurity, but they’re also out there in the world using all the platforms that the bad actors use in order to reach the same group of people,” said Michael Kaiser, president and chief executive of the nonprofit Defending Digital Campaigns. “That makes it extremely difficult to operate in this environment.”
The wider aim is to erode trust and sow confusion during and beyond election season. In July, the office of Senator Mitch McConnell released a photograph of him during a recent hospitalisation. Social media users immediately questioned whether it was genuine, with many suggesting the image was old or edited using AI tools. Digital forensics experts found no evidence that the photo had been altered or was dated. Ford noted that manipulating an image once demanded significant time and skill, whereas generative AI now allows a photo to be copied into an application and edited using plain-language instructions.
