What is AI, and how could it put elections and democracy at risk?
LWV Chicago’s Chicago in Focus lecture series recently featured a talk from Kevin B. Boyd, Chief Information Officer at the University of Chicago and Matt Morton, Chief Information Security Officer at the University of Chicago, about the potential positive and negative impacts of AI on our elections. The lecture was very entertaining and informative, with speakers who skillfully combined expertise with an engaging and interactive presentation style. The speakers' enthusiasm and sense of fun made the content more accessible—and listen for the surprise around the 10-minute mark!
Mr. Boyd and Mr. Morten explained that generative AI works by analyzing information in its “corpus” to generate new content based on probability of the proximity of certain words. For example, if a user asks an AI engine to write a cake recipe, it will scan through all the cake recipes in its corpus to create an entirely new recipe, based on the words that are likely to be found in other cake recipes.
Potential Positive Impacts
Enhanced campaign strategies: AI can analyze voter data to tailor messages, improving voter engagement.
Voter education: AI can generate informative content, making complex issues more accessible to voters. For example, a voter could ask AI to summarize the backgrounds and political positions of candidates running in a local election.
Potential Risks and Challenges
Deepfakes: Manipulated audio/video of candidates can harm reputations and influence voter perceptions.
Misinformation: Manipulated A/V content can spread misinformation, leading to confusion and mistrust among the electorate.
Character assassination: Politically-motivated manipulated A/V content can be used to tarnish the reputation of public figures.
Influence on voter perception: Voters may develop strong opinions based on visual content, which is often perceived as more credible than text.
Polarization: By creating or exacerbating divisions, manipulated A/V content can deepen societal and political divides.
Election interference: Hostile entities, both domestic and foreign, can use manipulated A/V content to interfere in elections, aiming to destabilize the democratic process.
What can we do?
Digital literacy education: Educate the electorate about the prevalence of manipulated media.
Fact-checking and verification: Promote and support fact-checking organizations that verify the authenticity of images and videos.
Regulation: Advocate for legislation that requires transparency in digital content.
Technology solutions: Support the development of AI-driven tools that can detect manipulated media.
Social media: Encourage social media companies to implement policies to identify and limit the spread of manipulated media.
Promoting ethical AI use: Encourage AI developers to adhere to ethical guidelines and incorporate watermarks or digital fingerprints in AI-generated content to indicate its artificial origins.
Public awareness campaigns: Inform voters about the potential misuse of AI in politics.
Watch the full lecture above and please share the video with your family and friends. You can also view the presentation slides.
This Chicago in Focus lecture was held Wednesday, November 8, 2023 at the University of Chicago Regenstein Library. The event was also presented virtually.
The event was cosponsored by UChiVotes, a student-led, nonpartisan voter engagement initiative founded out of the University of Chicago Institute of Politics to boost voter turnout and engagement on UChicago’s campus.
LWV Chicago's 'Chicago in Focus' lecture series seeks to highlight and explore the most important issues facing Chicagoans today.
