This article, written by Valerie Wirtschafter and published in 2024 on January 30th, discusses the potential threat of AI on democratic elections. AI would amplify the potential harm to the democratic process. Specifically, AI can construct realistic “deep fakes” that can spread misinformation to voters using the internet. This paper offers examples of how this has occurred in the past, drawing on elections from Slovakia and Argentina. In addition, Wirtschafter offers recommendations on how to protect information during the election year in the US – development, distribution, and detection as her main points. Development refers to the policy that monitors and or restricts the development of certain types of AI that interfere with elections – AI-generated images of politicians running for office, for example. If the technology is not built, it can not proliferate. Secondly, distribution refers to addressing and limiting how harmful generated content is spread. This area focuses on closing loopholes in manipulated media processes and increasing collaboration to better identify when harmful generated content is being spread to the public. Lastly, detection is the most vast area for information protection, it covers widespread education on misinformation, increased research on AI detection technology, and greater access to social media data. Overall, Wirtschafter’s article offers a condensed overview of the potential threat of AI on democratic processes and elections and discusses potential solutions to mitigate said risks.
