With the 2024 elections on the horizon, the use of artificial intelligence (AI) has sparked significant concern among lawmakers, especially as it pertains to the potential fabrication of deepfakes. At least half of U.S. states have initiated or enacted legislation aimed at regulating how AI technology is employed during electoral processes. This push, which has resulted from the growing anxiety over misleading information campaigns, sees both Democratic and Republican lawmakers coming together to formulate laws safeguarding the integrity of the electoral process.
California has taken the lead, with Governor Gavin Newsom recently signing three significant bills targeting the misuse of AI-generated content. The legislation aims to curb the spread of misinformation and disinformation amplified by deepfake technology, which can manipulate videos, images, and audio to deceive viewers. These laws build on earlier measures already adopted by states such as Texas and Michigan, which focused on protecting voters from AI-generated deceptions.
The circumstances leading to these regulatory moves were highlighted by incidents involving altered media being shared on social media platforms. For example, last summer, billionaire Elon Musk shared a deepfake video featuring altered audio of Vice President Kamala Harris, which went viral without any label indicating its misleading nature. This incident, along with others like it, has prompted urgent calls for clearer regulations on how AI-generated content can influence political campaigns.
According to analysis conducted by various news outlets, 26 states are now either considering or have passed legislation addressing the application of generative AI within electoral communications. Nineteen of those states have enacted laws explicitly concerning deepfakes, displaying growing bipartisan recognition of the risks posed by this technology. Texas, for example, was one of the first states to implement such measures back in 2019, prohibiting the creation or distribution of deepfake videos aimed at harming candidates or influencing voters.
This trend reflects broader public concerns about AI's potential to distort the electoral process. Experts have warned of the technology's capacity to confuse or suppress voters. Senator Mark Warner (D-Va.) emphasized the need for developers and platforms to acknowledge the misuse potential of generative AI, noting past instances where AI was employed to create impersonations aimed at misleading the electorate.
Legislation like the proposed NO FAKES Act, which seeks to protect individuals from unauthorized use of AI to create deceptive content, is gaining traction. Lawmakers from both parties have expressed support for this initiative, which has also drawn attention to the need for stronger legal frameworks governing AI technologies.
California's recent signing of new laws marks significant progress. Following Musk's viral deepfake incident, Newsom emphasized the necessity of maintaining electoral integrity and countering the manipulative power of AI. The newly signed Assembly Bill 2355 requires political advertisements to disclose whether they used generative AI to create or alter content, addressing the very concerns raised by recent events.
AB 2655, another piece of legislation, places liability on large online platforms, such as Meta and other social media giants, for failing to remove or properly label deceptive political content within 72 hours of user reporting. This signifies a shift toward increased accountability for tech companies, compelling them to act against misinformation spread via their platforms.
AB 2839, the third law signed by Newsom, prohibits the dissemination of misleading content about candidates during the 120 days leading up to and 60 days following elections if it is produced through AI. The laws set the stage for stringent repercussions against those who attempt to manipulate electoral outcomes using AI technology.
Overall, AI-generated misinformation poses deep threats to the electoral process, reminding the public of similar concerns raised during previous election cycles. The role of AI may offer some benefits for political campaigns, but many experts caution against the lack of comprehensive frameworks governing its use. Alix Fraser of the nonprofit Issue One stresses the need for flexible laws adaptable to the rapidly changing technology, potentially for the establishment of an agency dedicated to overseeing AI innovations.
Despite this growing patchwork of state regulations, many experts agree on the urgency of establishing federal legislation to uniformly address the intersections of AI and electoral integrity. Currently, there is limited federal oversight, with President Biden's AI executive order offering voluntary guidelines without enforcement power, leaving states to navigate the regulatory challenges individually.
The situation isn’t static, as states continue to explore their own personalized approaches to curb the influence of AI. For example, legislators from Pennsylvania, North Carolina, and Massachusetts are proposing bills requiring transparency around AI-generated content. Other states, like Delaware, have forwarded legislation focused on deepfakes but await gubernatorial approval.
Conversely, certain state initiatives have faltered; attempts by Alaska, Oklahoma, and Louisiana to regulate AI and deepfake technology have either been vetoed or stalled during committee reviews. Critics of these stalled proposals argue for urgency, contending AI's rapid development can outpace regulatory measures and extend opportunities for manipulation.
Legal experts and political analysts warn against merely addressing symptoms of the issue through temporary fixes. Many of the regulations focus on campaign ads, fail to recognize the broader, long-term risks of generative AI and deepfakes. Comprehensive and consistent federal guidelines are increasingly deemed necessary to combat the multifaceted challenges posed by AI and preserve the democratic process. Unless these systemic issues are addressed through well-defined federal statutes, many agree the nation might face continuous challenges with disinformation, potentially leading to voter manipulation and diminished trust.
Nonetheless, public awareness of AI impacts on politics has likely grown as some voters express skepticism toward both mainstream and alternative information sources. The influx of AI-created content adds to the confusion among audiences, contributing to the erosion of trust. Misinformation, spread through easily manipulated media, plants doubt across the political spectrum, with even verified content facing scrutiny.
Many agree the technology will continue to be used as both a tool for campaigning and as potential misinformation, paving the way for strategies likely becoming intertwined with public empathy strategies. To navigate these uncharted waters, many call for continued vigilance and proactive engagement across legal, technological, and political sectors to develop effective frameworks for managing the impacts of AI within the electoral system.
With the 2024 presidential race around the corner, states continue to grapple with AI's growing role, actively seeking solutions to manage its influence on voting behavior and public perception. The emergence of regulation efforts offers hope, but it remains clear the path forward requires commitment and consistent collaboration between lawmakers, technology platforms, and civic society aimed at safeguarding the democratic process.
