Artificial intelligence (AI) is already having a significant impact on electoral processes around the world, and the United States is no exception. Technological progress opens up new opportunities for analytics, communication with voters, and campaign management. On the one hand, AI allows politicians and campaigns to more accurately predict results based on big data and behavioral analysis. On the other hand, there is a threat of AI being used for disinformation, manipulation of public opinion, and even cyberattacks aimed at undermining trust in the electoral system. The US elections are a critical moment when the use of AI can become both a tool for positive change and a source of new challenges for democracy.
AI is a danger for democracy
The 2024 U.S. elections will be the first presidential race in the country to take place in the era of rapid development and spread of AI. Manipulations in the information space have jeopardized the transparency and objectivity of the 2016 and 2020 elections. The impact of much more adaptive and advanced AI technologies in 2024 may be unprecedented, creating new challenges for the electoral process.
Discussions about the threat of AI once again shook the United States in early 2024, when representatives of the Democratic Party endorsed their candidate for the presidential election. Although the US has a tradition of nominating the current head of state for a second term, the Democrats had to go through the election procedure before nominating Joe Biden.
At the same time, an incident occurred in New Hampshire: thousands of local residents received a call asking them not to vote in the next day’s primaries. The thing is, the voice on the other end of the phone sounded like President Joe Biden-though it wasn’t him.
Biden predictably won that primary, but concerns about the risks of AI grew. It turned out that a political consultant was behind the development of the diplomatic fake, and he said that he was only trying to warn about the dangers of using AI, not to influence the outcome.
Another scandalous example was the imitation of Kamala Harris’s voice in a parody election video created by blogger Mr. Reagan: in it, the Democratic candidate allegedly told typical gossip about herself that is usually spread by Republicans. The story gained even more traction online after Elon Musk shared it. For three days, the influential billionaire and owner of social network X did not clarify that the video was a parody, so many people took it seriously.
Donald Trump’s skillful political strategists were even able to use the topic of fighting fake news to their advantage, proving that sometimes you can simply accuse a political opponent of doing what you do. A Republican candidate accused Kamala Harris of “painting” the crowd in photos from her rally in Detroit. An expert review did not reveal the use of AI tools in the Democratic camp, but Trump successfully sowed doubts about the integrity of the Harris campaign.
At the same time, Trump himself often resorts to spreading fake news on social media. For example, the politician prematurely claimed to have received the support of world-famous singer Taylor Swift. Back in August, he posted a photo of the superstar’s fans drinking coffee in T-shirts with the words “Swiftie for Trump.” When Swift herself endorsed Kamala Harris in September, Trump’s team preferred to ignore it.
Is there any control over AI?
Candidates can be accused of using fake content in their campaigns. However, what to do when they have nothing to do with it – the content is simply distributed by the candidates’ supporters, which quickly makes the content viral. For example, in March 2024, Trump supporters distributed dozens of generated photos showing the Republican candidate with black voters. According to BBC Panorama, there is no evidence to directly link the fakes to the Trump campaign.
The wide availability of AI image generators and chatbots only increases the risks of such manipulations. There is no consensus among AI generator companies on what safeguards should be in place. For example, OpenAI, the developer of DALL-E and ChatGPT, prohibits users from creating images of public figures, including political candidates. At the same time, Grok, a new AI image generator on social network X, has fewer restrictions.
Not only the generated photos and dipshots pose a threat, but also the political position of the AI. For example, Grok, according to a study by Global Witness, when asked to create an unbiased post, offers content with explicit or implicit support for a particular political force. It also spreads false information about voting deadlines. However, not all chatbots react in this way. For example, Gemini refuses to answer political questions and redirects users to Google.
The lack of legislative regulation of the technology at the federal level also contributes to the permissiveness of AI. At the level of individual states, California, Texas, Minnesota, Michigan, and Washington, D.C., have successfully implemented laws restricting the use of AI during elections. However, federal legislative progress remains slow. Proposals in the US Senate and House of Representatives aimed at regulating AI in political campaigns have not moved forward.
This year, the Federal Election Commission and the Federal Communications Commission have just started discussing who has the authority to regulate the use of AI in political advertising. In early August 2024, in an interview with the Wall Street Journal, FEC Chairman Sean Cooksey stated that the agency has neither congressional authority nor technical expertise to regulate the use of AI in political campaigns. Thus, it is likely that unified rules will not appear before the November 2024 elections.
The impact of AI on the US elections is twofold: on the one hand, this technology opens up new horizons for increasing the efficiency of election campaigns, analyzing voter behavior, and improving communication with citizens. On the other hand, the use of AI poses serious risks associated with disinformation, manipulation, and data breaches. Therefore, it is important that the government, tech companies, and society in general develop an ethical and legal framework to regulate the use of AI in electoral processes. This will help ensure fairness and transparency of elections, while harnessing the benefits of technology to strengthen democracy.
artificial intelligence / copyright / election / intellectual property
