By LEAGUE OF WOMEN VOTERS MINNEAPOLIS
In the evolving landscape of democracy, elections have long stood as the cornerstone of citizen participation and political representation. But with the rise of data-driven campaigning powered by artificial intelligence (AI) technology, the electoral process is rapidly being revolutionized in unprecedented and unpredictable ways. While AI represents clear progress for campaign outreach, messaging, and fundraising, all benefitting the electoral process, it also poses a threat to the very foundational principles of democratic systems by enabling the widespread dissemination of disinformation. Here we take a closer look at the enormous impact of these AI transformations on campaigns and elections worldwide – and what can be done about it.
What is Artificial Intelligence (AI)?
Artificial Intelligence (AI) AI is advanced computer technology that uses smart, human-like cognitive abilities to analyze enormous amounts of data in order to perform tasks such as making predictions and recommendations or to look for solutions to problems. Many AI applications have become part of our daily lives: virtual assistants like Siri and Alexa, personalized content suggestions on Spotify and Netflix, Face ID on smartphones, and online search engines such as Google and Amazon.
While AI has been around for decades, its potential and influence increased significantly when ChatGPT, a powerful language processing technology that generates human-like text responses to user prompts, was introduced in November 2022. ChatGPT has been fine-tuned to engage in dialogue and simulate conversations more effectively than ever before. Its emergence has vastly expanded the possibilities and accessibility of generative AI (AI that generates text, images, videos) to millions of users worldwide through apps and consumer software. It has quickly become the fastest-growing consumer application in history.
Benefits and Dangers of (AI)
Communications and messaging in elections have greatly benefited from data-driven campaigning. Political parties and candidates can leverage AI-powered tools, including ChatGPT, to identify key issues and understand voter sentiment, resulting in improved personalized voter outreach, reduced administrative costs, and potentially higher voter turnout and a better-informed electorate.
However, the risks of AI misuse in elections currently far outweigh the benefits. AI’s powerful ability to supercharge disinformation by spreading deceptive and misleading content on a large scale with the intent to harm poses a significant threat.
One imminent danger is the newly created phenomenon of deepfake images and videos. These highly realistic digital manipulations, created by AI tools, can be used to promote election disinformation and produced by anyone using readily available technology in just a matter of minutes. Deepfakes are typically generated to replicate the appearance and voice of specific individuals without their consent. They are often so convincing that they can be difficult to distinguish from real photos or videos. FBI Director Christopher Wray recently warned about the growing threat, saying generative AI makes it easy for “foreign adversaries to engage in malign influence.”
Another risk is the ability to manipulate election outcomes by discrediting candidates or misleading voters through targeted biased campaign messaging on a much larger scale than ever before. One tactic involves producing multiple deepfakes loaded with factual inaccuracies and misrepresentations of information; another is inundating Facebook and online forums via accounts posing as real users with fake information generated by AI tools. This can significantly influence voter opinions and decisions, and potentially alter election results.
How to Recognize Fake Information
Often, it’s hard to know if something is real or not. But there’s good news: taking personal responsibility for critically evaluating information can be effective. If something seems off, try to exercise some caution—things aren’t always as they appear. Ask questions before drawing any conclusions. It’s especially important to avoid sharing information that might be false and intentionally malicious.
The National Conference on Citizenship has developed a useful tool designed to help detect fakes which is referred to as SIFT: Stop, Investigate the claim, Find better sources, and Trace the claim to its original source. While not foolproof, examining the content and identifying its original source can often provide insight into its intent and help determine its trustworthiness.
Addressing the Dangers of AI in Minnesota Elections
In 2023, a new Minnesota State Law was enacted that bans election-related deepfakes. It makes it a crime to share a deepfake within 90 days of an election if the person knows, or should have known, it was a deepfake and acts with the intent to harm a candidate’s reputation or influence the election result. The office of the MN Secretary of State is training and supporting election administrators to ensure they are aware of new technologies and legal changes.
MN Secretary of State Steve Simon’s office is also collaborating with tech partners like Open AI, Microsoft, Meta, Snap, and Google through the National Association of Secretaries of State. They share resources and get feedback on using their platforms to direct people to trusted sources.
Public messaging is another critical component of their strategy. Secretary Simon and their office members frequently engage with news media, voter outreach groups, and the public to educate about AI and its implications. The MN Secretary of State website (sos.state.mn.us) is continually updated and serves as a resource for Minnesotans to understand how the voting system works and how to be smart media consumers.
On a federal level, Senators Amy Klobuchar and Susan Collins support bipartisan measures to address AI’s impact on elections at the national level highlighted here. Secretary Simon recently testified in the US Senate Committee hearing on AI and election, calling for a multilevel approach to the challenges posed by emerging AI technologies and advocating for the federal government and the private sector to work alongside state governments to prevent the creation and amplification of dangerous election disinformation.
For reliable information on elections, visit trusted sources like: MNvotes.gov/elections-voting, Vote411.org, and Minneapolis Elections & Voter Services. Staying informed and critical of the information is essential in navigating the AI revolution in the electoral process.
This is the third article in the League of Women Voters Minneapolis 2024 Democracy Series. The next article, Pledge to Vote: Nuts & Bolts, will provide practical information to assist you in voting. All articles will be available at lwvmpls.org/for-voters.
