The 2024 US Election: AI Disinformation’s Looming Threat
The 2024 US presidential election has been a tumultuous race, marked by a last-minute candidate switch for the Democrats, assassination attempts against the Republican candidate, and a remarkably close contest according to polls. While many predicted a surge of AI-generated disinformation, the reality has been more nuanced. Although AI has amplified the volume and speed of false narratives, the core tactics of disinformation actors remain largely unchanged, relying heavily on established social media platforms. These platforms, unfortunately, seem more preoccupied with developing profitable AI tools than addressing the erosion of trust caused by the spread of misinformation.
While the anticipated "deepfake apocalypse" hasn’t materialized, "cheap fakes"—manipulated images, out-of-context content, and emotionally charged headlines—continue to effectively spread disinformation. Democratic candidate Kamala Harris has been a particular target, with AI-generated images depicting her in various false and compromising situations. Similarly, fabricated images of Republican candidate Donald Trump have circulated, showcasing him in unrealistic scenarios. Even the vice-presidential candidates have not been immune, with AI-generated content targeting both Tim Walz and JD Vance.
A significant concern remains the potential for AI-generated audio disinformation to spread via WhatsApp, a platform heavily used by Hispanic and Latino communities for political discussion. While the feared AI-driven disinformation revolution hasn’t yet occurred, the election cycle has witnessed the resurgence of ten recurring disinformation narratives identified in previous elections across Latin America and North America. These narratives range from unsubstantiated claims of electoral fraud to misinformation about voter eligibility and voting procedures.
The most prominent disinformation narrative in 2024 revolves around voter eligibility, specifically the false claim that ineligible individuals, particularly undocumented immigrants, are voting or will vote. This narrative, often amplified by Republican candidates and their supporters, has gained significant traction online, sometimes boosted by paid advertisements on platforms like Facebook and Google. This echoes the “Big Lie” narrative promoted by Trump in 2020, which was repeatedly debunked by over 60 federal courts.
To counter the pervasive disinformation and empower voters, Factchequeado has developed two innovative tools: Electopedia and Electobot. Electopedia, a bilingual website, provides comprehensive information about the US election process, including voter registration, voting methods, and polling locations. It also offers educational resources to inoculate communities against disinformation and encourage voter participation. Electobot, a Spanish-language chatbot on WhatsApp, provides real-time answers to election-related questions, combats misinformation, and offers location-based voting information.
These initiatives represent a proactive approach to combatting disinformation, recognizing the specific vulnerabilities of Spanish-speaking communities. By providing accurate and accessible information, Factchequeado aims to empower voters to make informed decisions and resist manipulation. As the election draws closer, the threat of AI-powered disinformation remains a serious concern. While the full potential of this technology for malicious purposes has yet to be realized, the existing tactics continue to erode trust and undermine democratic processes. Vigilance and proactive measures like those implemented by Factchequeado are crucial in navigating the remaining days of this critical election cycle. The stakes are high, and the fight against misinformation demands constant attention and innovative solutions.