The use of artificial intelligence has raised major concerns for employees, the technology industry, and now politics. A new political ad is using images created by AI to paint a dystopian picture of the future under President Biden, raising concerns about misleading information born of machine learning. 

In response to President Biden’s reelection announcement on Tuesday, the Republican National Committee released a digital ad showing what they think would happen to America should the president win in 2024. 

The ad contains a number of AI-generated images, including one of Mr. Biden and Vice President Harris celebrating their 2024 victory. Other, darker images include boarded-up banks, tens of thousands of migrants crowded at the southern border, global unrest in the wake of Communist China’s invasion of the Republic of China on Taiwan, and another image showing martial law in San Francisco. 

The ad was first reported on by Axios’ Alex Thompson.

The executive director of the Democratic National Committee, Sam Cornale, responded to the ad on Tuesday, saying AI is now being used to replace political consultants. “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,” Mr. Cornale wrote. 

Beyond creating misleading images, AI has also been proven to be successful in recreating the voices of politicians. One new software allows users to insert scripts and choose a voice to read it. A popular trend on the social media app TikTok shows Presidents Biden, Trump, and Obama playing video games together. A series of videos showing AI-generated versions of the men have received tens of millions of views in recent months, in part due to the convincing scripts and voice recreations.

Realistic images or videos combined with these scripts raise concerns about so-called “deepfakes” in American politics, which is the use of AI to create convincing videos of government officials or well-known persons in order to paint them in a bad light. 

Two senators have been raising concerns about these deepfakes for years. Senators Rubio and Warner — the top legislators on the intelligence committee — have introduced legislation so that the defense and intelligence communities can better deploy AI for safety purposes, but still harbor concerns about the technology’s use in everyday life. 

Mr. Rubio said as far back as 2018 that AI could be used to disrupt American political culture. “Imagine a video like that delivered on the eve of an election in a culture who has a bias toward believing outrageous things, a media that is quick to promulgate it and social media, where its spread can’t be stopped,” he said at a committee hearing. 

AI “could be used to produce fake videos of American politicians using racial epithets or taking a bribe,” Mr. Rubio added. “It could be used to show a U.S. soldier massacring civilians overseas or a U.S. official admitting a secret plan to some conspiracy theory.”

Little progress has been made in stopping potential deepfake ads at the federal level, but some state governments are looking to take the lead. The governor of Washington, Jay Inslee, will soon sign a piece of legislation that provides legal recourse for political candidates who have deepfake videos deployed against them during campaigns. 

Washington’s secretary of state, Steve Hobbs, told ABC News that he was “trying to get ahead” of these videos. “The last thing you want is political campaigns or political action committees to put ads out of the deepfakes of the person they are trying to vote out,” he said. “Even if that ad was up a day or two before people realize it’s a deepfake it can cause damage.”