In the spring of 2023, a notable experiment in the integration of artificial intelligence into the political realm took place in Wyoming, USA, with the candidacy of a virtual integrated citizen (VIC), an AI bot based on ChatGPT. This initiative was led by human creator Victor Miller, who aimed to explore how AI could theoretically perform governance functions. His campaign, while not resulting in a victory, signalled a unique intersection between technology and democratic processes.
As the year progressed, many analysts speculated on the potential impacts of generative AI on democratic elections, especially with over 2 billion individuals participating in electoral processes across more than 60 nations in 2024. Initial assessments indicated that generative AI might play a pivotal role, raising alarms about both its capabilities and the risks it posed to election integrity. However, recent expert analyses suggest a reevaluation of these early predictions. Conversations with industry specialists have indicated that generative AI likely had little to no significant effect on these elections, marking a turn in the narrative surrounding the anticipated "AI election."
A central concern surrounding the use of generative AI in elections was the proliferation of deepfakes—manipulated media that could potentially mislead voters. Scott Brennen, director of the Center for Technology Policy at New York University, noted, “I think concern about misleading deepfakes was taking up a lot of oxygen in the room,” referring to the widespread fear of AI-generated content that could distort public perception. Nevertheless, despite the apprehension, many campaigns exhibited a cautious approach toward the use of deepfake technology, as its complexity deterred some from leveraging it for political gain.
In the United States, there was a prevailing apprehension about navigating an evolving landscape of state-level legislation aimed at combating deceptive AI practices. Brennen added, “I don't think that any campaign or politician or advertiser wants to be a test case, particularly because the way these laws are written, it’s sort of unclear what ‘deceptive’ means." This caution underscores a significant barrier to the adoption of generative AI in political advertising and campaign strategies.
Earlier in the year, WIRED initiated the AI Elections Project, intended to monitor the role of AI in elections globally. An analysis conducted by the Knight First Amendment Institute at Columbia University revealed that approximately half of the reported instances of deepfakes were not designed with deceptive intent. This observation supports findings from The Washington Post, which indicated that while deepfakes may not have effectively misled the public or swayed opinion, they contributed to heightened partisan divisions within the electorate.
As the landscape of political campaigning continues to evolve, the interplay between AI and democratic institutions remains a complex and ever-changing narrative. While the anticipated impacts of generative AI did not unfold as dramatically as once envisioned, its presence in political discourse and the caution surrounding its application suggest an ongoing dialogue about the future role of technology in governance and electoral integrity.
Source: Noah Wire Services