By: Avi Samy

Generative AI

Generative AI, an advancement in artificial intelligence (AI), represents a leap in machine learning capabilities. Unlike traditional AI models which solve specific tasks while relying on predefined rules, generative AI algorithms generate new content. Examples of traditional AI include voice assistants like Siri or Alexa and even recommendation engines on websites and other platforms. Generative AI can create stories, images, emails, etc. based on a provided prompt. Trained on vast datasets, generative AI can learn complex patterns and relationships (Martineau, 2023). Companies like OpenAI, Google, and Bing utilize generative AI (see Figure 1). Leveraging this knowledge, generative AI creates realistic content in the form of text, images, and videos, resembling human-generated material. This innovation offers possibilities for creative expression, content generation, and problem-solving in many different fields. However, its proliferation raises ethical and societal concerns, especially regarding potential misuse in election campaigns, threatening electoral integrity. Robust safeguards and ethical guidelines are urgently needed to protect democratic principles and public trust.

Figure 1

Some of the biggest Generative AI Companies

Source: IOT

A Watershed Election Year

The year 2024 emerges as a pivotal election year for the world, marked by unprecedented global challenges and transformative shifts in geopolitical dynamics. About 70 countries, which account for nearly half the world’s total population, are preparing to have national elections this year (Simon, 2024). As shown in Figure 2 below, an expected 2 billion people are expected to vote, including in significant regions such as the United States, India, and much of Europe. Against the backdrop of climate change, debates on human rights, and ongoing wars, the outcomes of elections in key nations will significantly shape the trajectory of international relations, economic policies, and collective efforts to address pressing global issues. The choices made by voters and leaders in the 2024 elections carry profound implications for the future of democracy, human rights, and global stability.

Figure 2

2024 Election Statistics

Source: Statista

Unveiling the Deception: AI’s Misuse

US President Joe Biden’s voice was heard across the state of New Hampshire, just days before the state’s presidential primary. Campaigning efforts were beginning to ramp up as candidates looked to cement their cases with voters across the states. There was just one problem: the President was instructing people not to bother voting. After a security analysis, a definitive conclusion was reached. It wasn’t the President speaking. After listening to the recording a few times, it was quite obvious that the voice was generated by AI (Mcmillan, 2024).

Similar instances were seen in 2023, with a fake video of Biden declaring a national draft to aid Ukraine’s war effort as well as a deepfake depicting Senator Elizabeth Warren claiming that Republican voters should be barred from voting. Recent advancements in technology have resulted in images and videos that appear real, online bots that pose as humans, and several other problems. Social media companies currently have their hands full, addressing questions on their content, cyberbullying and harassment, the spread of misinformation, online safety, and a host of other issues. Add AI’s role in politics and society to that growing list, and it is clear that there is a severe issue at hand. The corporations claim that policies are in place to prevent the use of AI in deceptive ways, but the extent to which they are enforced remains unclear (Ingram, 2024).

Figure 3

Despite the debunking of the robocall, many remain concerned about the security of elections

Source: Wired

An Emerging Issue

 AI could synthesize pictures of people who haven’t existed for years, and now a single prompt can be used to generate multiple realistic images. Deepfake audio (the use of artificial intelligence to create convincing sentences of specific people has also grown significantly as AI now requires less training to clone one’s voice (Panditharatne, 2023). Social media companies have grappled with questions on political content for years. Strict monitoring of political discourse online was apparent in 2020 in response to claims about Russian interference in what was the upcoming presidential election. Fears of a potential repeat of 2016 where Russians employed hundreds and possessed millions of dollars in an attempt to ensure a victory for then-candidate Donald Trump (Mcmillan, 2024). In a survey of around 1,000 voters, 70% believed that Russia interfered, as seen in Figure 4.

In recent years companies have loosened many restrictions that were once in place. After Elon Musk acquired Twitter (now called “X”) in 2022, many suspended accounts were reinstated and many workers were laid off due to the company’s financial issues. YouTube halted their removal of videos claiming fraud was present in past elections, citing that restrictions on political speech would lead to new concerns. In addition, Meta allowed political ads questioning the legitimacy of Biden’s 2020 victory over Trump to run on their platforms. 

Figure 4

The majority of people from each political party believed interference was seen in 2016

Source: PBS

Countering AI Manipulation

Since the creation of ChatGPT, Bard, and other generative AI models, our ways of obtaining information have changed dramatically. AI has been used to produce articles at faster rates and lower prices, despite the tendency for it to produce misinformation (Panditharatne, 2023). Efforts to fight this and AI’s impact on politics are appearing, as several weeks ago twenty tech companies utilizing AI signed a pact to prevent their technology from impacting the upcoming election (Ingram, 2024). This includes tech giants such as Microsoft, OpenAI, Google, and Meta, some of which are seen in Figure 5. The main reason for the creation of this misinformation is due to the lack of regulation and action taken by lawmakers. Still, many are left unsatisfied that these companies didn’t place a full ban on AI having the ability to impact election content. Meta has stated that it would attempt to label AI-made images, although several complications exist when trying to do this, especially when considering the inclusion of audio and video (Ingram, 2024).

Figure 5

Amazon, Apple, Meta, and Google are part of the accord to combat AI

Source: Slate

Future Implications

The advent of generative AI presents profound challenges, particularly in the realm of electoral politics. As the world approaches a crucial election year in 2024, the misuse of AI-generated content poses a significant threat to the integrity and fairness of democratic processes worldwide. While efforts to combat this threat are underway, including initiatives by tech companies to self-regulate AI usage, the lack of comprehensive regulation and enforcement remains a pressing concern. As society grapples with the implications of AI in politics and society, it becomes increasingly imperative to establish robust safeguards and ethical guidelines to mitigate the risks and uphold the principles of democracy. Only through collective action and responsible stewardship of AI technology can we safeguard the integrity of elections and preserve public trust in democratic institutions for generations to come.

References and Sources

Ingram, D. (2024, February 16). Microsoft, Google and Meta pledge to prevent AI election interference. NBC News. Retrieved March 7, 2024, from https://www.nbcnews.com/tech/tech-news/microsoft-google-meta-pledge-action-ai-election-risks-rcna138909

Martineau, K. (2024, February 14). What is generative AI? IBM Research Blog. Retrieved March 7, 2024, from https://research.ibm.com/blog/what-is-generative-AI

McMillan, R. (2024, February 15). New Era of AI deepfakes complicates 2024 elections. New Era of AI Deepfakes Complicates 2024 Elections. https://www.wsj.com/tech/ai/new-era-of-ai-deepfakes-complicates-2024-elections-aa529b9e 

Panditharatne, M. (2023, June 13). How AI puts elections at risk — and the needed safeguards. Brennan Center for Justice. Retrieved March 7, 2024, from https://www.brennancenter.org/our-work/analysis-opinion/how-ai-puts-elections-risk-and-needed-safeguards?gclid=Cj0KCQjw-pyqBhDmARIsAKd9XIM7iPmtI-3Y8V22nZ4C-3dStTVRI8Xu1FhBKqlwBhaAyvmqGUFwuaIaAq4uEALw_wcB&ms=gad_ai%20voting%20system_665268346943_8628877148_150749630603

Simon, S. (2024, January 20). 2024 is a big election year around the globe. Will democracy win? NPR. Retrieved March 7, 2024, from https://www.npr.org/2024/01/20/1225795005/2024-is-a-big-election-year-around-the-globe-will-democracy-win