In June 2023, a manipulated video featuring Florida Governor Ron DeSantis circulated online as a fabricated endorsement advertisement. The deepfake, which used artificial intelligence to superimpose the governor's likeness and voice, was presented as a promotional spot for his presidential campaign. The video's authenticity was quickly challenged by fact-checkers and news organizations, who identified it as a synthetic creation rather than genuine campaign material. The incident drew attention to the growing role of generative AI in political communication and the challenges it poses for verifying electoral content.
The deepfake emerged during a period of heightened political activity, as DeSantis formally announced his candidacy for the 2024 Republican presidential nomination on May 24, 2023, via a Twitter Spaces event with Elon Musk. The fake ad, which appeared on social media platforms including Twitter (now X) and YouTube, depicted DeSantis in a manner inconsistent with his known public statements and campaign messaging. It was not produced by the official DeSantis campaign, which disavowed the video in statements to press outlets. The video's origin remained unclear, though analysts suggested it may have been created by an individual or group seeking to parody or discredit the governor.
Technical Characteristics
The deepfake employed techniques common in contemporary generative AI, including deep learning models trained on large datasets of facial images and audio recordings. Such systems, often built on neural network architectures like autoencoders or generative adversarial networks, can synthesize realistic video by mapping a target person's facial expressions and speech patterns onto a source actor. In this case, the video showed DeSantis speaking words he had not publicly uttered, with lip movements and vocal intonation that mimicked his mannerisms. Detection tools, such as those developed by researchers at institutions like MIT CSAIL and Stanford AI Lab, flagged inconsistencies in blinking, lighting, and audio frequency that indicated manipulation.
Platform Response
Social media platforms faced scrutiny over their handling of the deepfake. Twitter, under owner Elon Musk, had reduced content moderation staff in 2022, leading to concerns about the spread of misinformation. YouTube, owned by Google, maintained policies against manipulated content that could mislead viewers, but enforcement varied. Neither platform removed the video immediately, and it continued to circulate for several days before being flagged by third-party fact-checkers. The incident prompted calls for more robust detection and labeling systems, with some experts advocating for mandatory watermarks on AI-generated content.
Political and Legal Implications
The fake ad raised legal questions under U.S. election law, which prohibits fraudulent misrepresentation in federal campaigns. The Federal Election Commission had not yet issued specific rules on deepfakes, though some states, including California and Texas, had enacted laws criminalizing deceptive AI-generated political content. Florida itself had no such statute at the time, leaving a regulatory gap. The DeSantis campaign did not pursue legal action, but the episode contributed to broader debates about the integrity of the 2024 election cycle.
Broader Context
This incident was part of a wave of AI-generated political content in 2023. Earlier that year, the Republican National Committee released a purely AI-generated ad depicting a dystopian future under President Joe Biden, which was clearly labeled as synthetic. In contrast, the DeSantis deepfake was not labeled, making it more deceptive. Researchers at BAIR (Berkeley AI Research) and University of Oxford published studies in mid-2023 showing that deepfake detection remained unreliable, with error rates above 10% even for advanced systems. The episode underscored the need for media literacy and technical safeguards as generative AI tools became more accessible.
Aftermath
Following the incident, several advocacy groups, including the nonprofit OpenPanel, called for federal legislation to regulate AI in political ads. In August 2023, the Federal Election Commission announced it would consider a petition to clarify its authority over deepfakes, though no rule change occurred before the end of the year. The DeSantis campaign itself did not publicly address the video beyond a brief denial, focusing instead on traditional campaign events. The deepfake remained a case study in academic and policy discussions, cited in reports by Carnegie Mellon University and Stanford AI Lab as an example of emerging threats to democratic processes.