By: Nathaniel J. Greene, Community & Culture Reporter

Artificial intelligence (AI) is rapidly transforming numerous aspects of our lives, from everyday conveniences to complex societal systems. However, its impact on the electoral process is raising significant concerns. At a recent Ethnic Media Services briefing titled “Artificial Intelligence and the 2024 Elections,” experts underscored the profound implications of AI on the integrity of election systems. Held on July 12, the briefing featured insights from Jonathan Mehta Stein, Executive Director of California Common Cause; Jinxia Niu, Program Manager for Chinese Digital Engagement at Chinese for Affirmative Action; and Brandon Silverman, former CEO and Co-Founder of CrowdTangle.

AI’s Role in Disinformation Campaigns

Artificial intelligence has become a powerful tool for spreading disinformation, a threat that is becoming increasingly sophisticated. As Stein pointed out, “The power of AI to generate synthetic content that can mislead voters and stoke disinformation is unprecedented. What used to require a studio budget and a production team can now be accomplished with a few clicks.”

Stein highlighted a recent case where the U.S. Department of Justice disrupted a Russian disinformation campaign involving thousands of fake social media profiles, many generated using AI. These profiles, posing as regular Americans, pushed pro-Russia propaganda, illustrating the scale and ease with which AI can be weaponized in disinformation efforts.

California Common Cause has responded to these threats by launching CITED (California Initiative for Technology and Democracy), which aims to combat online disinformation and other AI-related electoral threats.

Stein emphasized the urgency of addressing these challenges, stating, “We’re now entering the first AI election in which deep fakes and AI disinformation have the power to inundate our political discourse and create images, audio, and video that voters don’t know if they can trust.”

The Danger to Voter Trust

The potential for AI to undermine trust in the electoral process is a significant concern. Stein explained, “Imagine false audio of a county elections official caught on tape saying their voting machines have been hacked, or a fake robocall from Governor Newsom going out to millions of Californians on the eve of Election Day telling them their voting locations have changed. These threats are not theoretical.”

He noted that similar incidents have already affected elections in countries like Bangladesh, Slovakia, Argentina, Pakistan, and India.

Local elections are particularly vulnerable. Stein warned, “If there is a deep fake of President Biden, it will only take a few hours before the national press exposes it as fake. But if there’s a deep fake of your city council member in handcuffs with a fake news story that they were just arrested for taking a bribe, there might not be as much media coverage.”

Challenges in the Chinese American Community

Jinxia Niu from Chinese for Affirmative Action shared insights on how AI-powered disinformation is impacting the Chinese American community. Niu’s organization runs a Chinese language fact-checking website, and their latest report has documented over 600 pieces of disinformation in the last 12 months.

“We are seeing AI-generated photos and videos both supporting Trump and attacking Biden, being widely shared by Chinese social media influencers,” Niu explained.

One major challenge is the lack of timely, effective, and accurate information in the immigrant community. Niu stressed, “Our immigrant community media is so limited and outnumbered. Many AI-generated disinformation pieces go unchecked because there are not enough resources to address them.”

She also highlighted the difficulty in combating disinformation on encrypted messaging apps like WeChat and Telegram, which are popular among Chinese Americans.

“These platforms create a perfect dilemma, making it so difficult, if not impossible, to intervene with disinformation,” Niu said.

The Ethical Responsibility of Social Media Platforms

Brandon Silverman, former CEO of CrowdTangle, discussed the role of social media platforms in amplifying AI-generated disinformation. He emphasized that most misinformation does not violate platform policies because it falls into a gray area of being misleading but not outright false.

Silverman explained, “It’s much harder to build a system that can track political disinformation. Platforms can easily hunt for and flag sexual exploitation or racist slurs, but the sophisticated nuance of political disinformation makes it much more challenging.”

Silverman also addressed the disproportionate impact of social media harms on communities of color.

“It is often elite and trusted sources in media information ecosystems who are the most important spreaders of disinformation,” he said.

This reality underscores the need for better policies and more resources to support local and ethnic media.

Solutions and the Path Forward

To combat these threats, Stein proposed digital watermarking as a legislative solution.

“Mandating generative AI companies to embed provenance metadata in all digital media they create would help identify AI-generated content. Social media platforms could then label such content appropriately,” Stein explained.

This approach aims to provide transparency and help voters distinguish between authentic and AI-generated content.

Niu emphasized the importance of training and equipping fact-checking teams within immigrant communities.

“We need to train our staff quickly to understand how AI works and how to debunk AI-generated disinformation,” she said.

This capacity-building is crucial for maintaining the integrity of information within these communities.

Silverman suggested focusing on consistent narratives rather than individual pieces of content.

“Identifying and pushing back against false narratives is more effective than responding to each misleading post,” he advised.

He also highlighted the importance of collaboration and resource-sharing among organizations to tackle these challenges collectively.

Ensuring the integrity of the electoral process requires a concerted effort from policymakers, technology companies, media organizations, and the public. By staying vigilant and informed, we can navigate this new landscape and protect the democratic process from the threats posed by AI-generated disinformation.