Republicans Launch Eerie AI-Generated Attack Ad On Biden

Topline

The Republican National Committee debuted an apocalyptic foreshadowing of a second Biden-Harris term moments after President Joe Biden announced his re-election campaign on Tuesday in a video rebuttal generated entirely by artificial intelligence—the latest sign computer-generated content could shift the political campaign landscape.

U.S. President Joe Biden and Vice President Kamala Harris on the South Lawn of the White House on … [+] April 08, 2022 in Washington, DC. (Photo by Anna Moneymaker/Getty Images)

Getty Images

Key Facts

The video shows AI-generated images of Biden and Vice President Kamala Harris winning on election night next year, followed by simulated scenes of explosions in Taiwan, police in tactical gear lining the streets of San Francisco, migrants flooding the southern border and Wall Street buildings abandoned.

Predictive newscasts narrate in the background: “Who’s in charge here? It feels like the train is coming off the tracks.”

An RNC spokesperson told Axios the video was produced entirely through artificial intelligence, which Democrats and Republicans have both expressed plans to leverage during the 2024 election cycle in an effort to streamline operations, though the software has raised concerns about malicious fake content duping voters and campaign staffers.

Further Viewing

Chief Critic

Democratic National Committee Executive Director Sam Cornale suggested the RNC’s use of AI was an act of desperation in response to the video. “When your operative class has been decimated, and you’re following MAGA Republicans off a cliff, I suppose you have no choice but to ask AI to help,” he tweeted Tuesday.

Crucial Quote

RNC Chair Ronna McDaniel called Biden “so out-of-touch” in a statement warning “inflation will continue to skyrocket, crime rates will rise, more fentanyl will cross our open borders, children will continue to be left behind, and American families will be worse off” if he is re-elected.

Key Background

Campaigns reported using artificial intelligence to identify fundraising audiences during the midterm election, but in recent months, following the launch of OpenAI’s ChatGPT in November, campaigns have been testing chatbots to generate written content, including statements from candidates, ads and emails, the New York Times reported. But there are concerns that the software could be used to mislead the public via convincing voice messages that sound like candidates speaking directly to voters or chatbots posing as voters providing false information to campaigns, the Times noted. AI-generated parodies of prominent political figures have gone viral in recent months. Fake images of Trump being arrested and wearing prison garb were widely shared on social media ahead of Trump’s indictment in Manhattan Criminal Court. Trump also shared a faux image of himself kneeling in prayer on Truth Social in March.

Surprising Fact

While AI-generated content has entered a new era of legitimacy, the use of deepfake videos, audios and images has been around for years. A fake video of former President Barack Obama calling Trump a “total and complete dipshit” circulated in 2018, and in 2004, an altered image of then-Democratic presidential nominee John Kerry and actress Jane Fonda, depicting them at an anti-war protest in the 70s, was widely shared online.