AI’s Impact on the 2024 US Election: A Muted Roar
The 2024 US presidential election was widely anticipated to be the first significantly impacted by the rise of artificial intelligence, particularly in the realm of misinformation. Experts and policymakers raised alarms about the potential for AI-generated deepfakes to manipulate voters, influence election outcomes, and even destabilize the democratic process. These concerns prompted a flurry of legislative action and proactive measures by tech companies to mitigate potential harm. Sixteen states implemented laws regarding AI use in campaigns, often mandating disclaimers on synthetic media. The Election Assistance Commission published a toolkit to help election officials navigate the age of AI-fabricated information, and social media platforms implemented new policies to label or restrict AI-generated political content.
Despite these preemptive measures, the widespread, AI-driven election disruption that many feared failed to materialize. While misinformation played a prominent role in the 2024 election cycle, the primary vectors were familiar tactics: text-based social media posts, misleadingly edited videos, and out-of-context images. The anticipated flood of sophisticated deepfakes designed to sway voters simply did not occur. Instead, the misinformation landscape was dominated by traditional, non-AI methods. This unexpected outcome prompted reflection and analysis from experts in technology, policy, and misinformation.
Why did the AI-powered misinformation apocalypse fail to arrive? Several factors contributed to AI’s muted impact. Firstly, other, more readily available and effective methods of spreading misinformation remained dominant. Prominent figures with large followings could easily disseminate false narratives through traditional channels like speeches, media interviews, and social media posts. The reach and influence of these established methods diminished the need for more complex AI-generated content.
Secondly, the proactive steps taken by technology companies and policymakers appear to have played a crucial role in curbing the spread of AI-generated disinformation. Platforms like Meta, TikTok, and OpenAI implemented safeguards such as mandatory disclosures for AI-generated political ads, automatic labeling of synthetic content, and outright bans on the use of their tools for political campaigns. These actions, combined with state-level legislation, likely deterred some malicious actors and limited the potential for widespread AI-driven manipulation.
Thirdly, the AI-generated content that did emerge often served to amplify existing narratives rather than create entirely new, deceptive storylines. For instance, after false claims about Haitian immigrants circulated, AI-generated images and memes depicting animal abuse proliferated online, reinforcing the existing misinformation. This suggests that AI, in this context, acted more as an amplifier of existing biases and falsehoods than as a generator of entirely new deceptive narratives.
Despite the limited impact of AI-generated content in the 2024 election, its potential for harm was not entirely absent. Several instances demonstrated the capacity of AI to manipulate narratives and incite partisan animosity. The AI-generated robocall mimicking President Biden’s voice, designed to discourage voting in the New Hampshire primary, stands as a stark example. While isolated, this incident highlighted the potential for relatively simple AI tools to be used for malicious purposes.
Furthermore, while deepfakes did not become the dominant form of election misinformation, their presence was not negligible. Examples emerged of deepfakes being used to denigrate candidates or distort their views, often furthering pre-existing partisan narratives. Researchers found that deepfakes were primarily used for satire, reputational harm, and entertainment, but instances of politically motivated deepfakes were also documented. These instances, though not widespread, underscore the ongoing need for vigilance and development of effective countermeasures.
Looking beyond AI-generated content, foreign influence operations, another potential source of election interference, also largely relied on traditional tactics rather than AI. Intelligence agencies identified several foreign influence campaigns, but these typically involved actors in staged videos and other non-AI methods. While the potential for AI to be used in future foreign influence operations remains a concern, it did not play a significant role in the 2024 election.
The relative absence of widespread AI-generated misinformation in the 2024 election offers valuable lessons for future elections. The proactive measures taken by tech companies, policymakers, and researchers likely contributed to mitigating potential harm. However, the evolving nature of AI technology requires ongoing vigilance and adaptation. As AI tools become more sophisticated and accessible, the potential for misuse will likely increase, demanding continued efforts to develop effective safeguards and public awareness campaigns. The 2024 election, though not “the AI election” as many anticipated, served as a crucial testing ground and provided valuable insights into the evolving landscape of election integrity in the age of artificial intelligence.