It has become commonplace in US political campaigns. Ads that broadly advocate a dystopia if the opposing candidate wins. Manipulated, underexposed imagery and handpicked headlines combine to build a crescendo of doom.
But following Tuesday’s announcement that Democratic President Joe Biden is running for re-election, the official Republican video stood out for one particular reason. It was entirely generated using artificial intelligence (AI) images.
The Republican National Committee’s embrace of “the revolutionary technology of our time” isn’t surprising given the rapid progress and availability of AI products, said a senior at the Center for Innovation at the Brookings Institution. said researcher Darrell West.
The use of AI by Republicans is an early sign of what is to come, he told Al Jazeera.
“Three years ago, AI was underutilized in election campaigns. But the technology has advanced very quickly. I was. “The software he doesn’t have to be a designer or a video editing expert to create highly realistic videos.”
“It’s uncharted territory,” he added. “We’ve gone from just doing small parts of an image in Photoshop to generating a completely new image from nothing. will be able to project new realities that may not exist in
And then it starts – #generativeai It is used to create disinformation and deepfake election campaign videos.
The Republican National Committee has released an ad opposing Biden using an AI-generated image. Before the pandemic, we talked about the dangers of deepfakes… pic.twitter.com/7bw2yCn24Z
— Theo (@tprstly) April 26, 2023
“AI Generated Appearance”
The Republican National Committee has been transparent about its use of AI, a collective term for systems that attempt to mimic and surpass human cognitive skills such as learning, reasoning, and creativity.
In a YouTube description, the political group called the video “an AI-generated look at the potential future of this country if Joe Biden is re-elected in 2024.”
The video itself also included the text “built entirely with AI imagery.” Realistic images streamed as fake newsreaders announce Biden’s victory in 2024, followed by China’s invasion of Taiwan, financial market crash, southern border runaway, San Francisco shut down by officials Etc., “citing the rise of crime and the fentanyl crisis,” followed a series of fictional disasters. .
In many ways, the video represents a lesser departure from the imagery and rhetoric common in US campaigns.
As The Washington Post noted in 2020, “sharing defaced images of election rivals is an outdated strategy in modern politics”. It reported that the images “accelerated rapidly”.
Meanwhile, U.S. courts have repeatedly upheld broad interpretations of the right to make false or misleading election statements. Most recently, an appeals court ruled in his February that a North Carolina law prohibiting false campaigns was “likely unconstitutional.”
West, Technology Innovation Fellow at the Brookings Institution, said: In fact, a candidate can knowingly say something false and still be allowed to say it. ”
Tom Wheeler, who chaired the Federal Communications Commission under former President Barack Obama, said otherwise in an interview with NPR last year.
speed and sophistication
However, AI has the potential to enhance existing campaign practices, West said, adding that “there are virtually no checks on the use of this technology in the campaign environment.”
“There is no legal requirement to allow us to use images generated by AI,” he said. “In this case, the RNC voluntarily revealed it … but in the future there will be many organizations that will use it without notifying voters.”
So-called deepfakes (videos and audios that falsely portray what an individual says or does) are a source of particular concern, even though they have so far remained on the political line.
California and Texas passed legislation ahead of the 2020 general election. The former would allow misrepresented candidates to sue the creators of deepfakes, while the latter would impose criminal penalties on deepfakes.
However, federal efforts to enact legislation on the issue have made little progress, prompting questions about enforceability and calls from digital rights groups to argue that major technology platforms need to provide oversight. facing backlash.
‘Wild West’ moments
On the other hand, those seeking to influence campaigns “can react almost instantly” to the latest events, West said.
“Basically, we ask the AI to generate the images and you get them in a matter of seconds. may appear.”
“This is a very fast-paced campaign, with many allegations and counterclaims occurring every minute,” he said.
The New York Times reported in March that Democrats were also incorporating aspects of AI into their campaigns, testing the technology and writing the first drafts of several fundraising messages. .
The paper, citing three people familiar with the effort, said that the AI-generated and human-edited messages at the Democratic National Committee were “completely drafted by humans in terms of generating engagement and donations.” It was reported that the results were as good as or better than the original copy.
The ability to quickly reach and potentially misinform specific voter segments will be especially important in the 2024 presidential election, which West said could be “1-2% of voters.” is said to be high.
“This is a Wild West moment when it becomes impossible for voters to distinguish between the real and the fake,” he said.
“Things are going to come at them from all directions and there is a risk of widespread confusion,” he said. “And that can lead to bad decisions.”
