Experts Warn on the Dangers of AI-Based Disinformation
En una conferencia de prensa organizada por Ethnic Media Services el 12 de julio, un panel de expertos, incluidos Jonathan Mehta Stein, Jinxia Niu y Brandon Silverman, discutió la nueva frontera de la desinformación.The digital landscape is on the brink of a revolution. Generative artificial intelligence is becoming a stronger part of our everyday lives and is now integrated with Meta platforms like Instagram and WhatsApp and will soon be part of Apple’s iOS. With these changes, the way we interact with content is undergoing a rapid transformation. Feeds are now flooded with artificially generated texts, images, and audio, representing a new frontier for misinformation. This shift is already impacting politics, and all audiences need to prepare to navigate digital information amidst this influx of artificial content.
At a press conference by Ethnic Media Services on July 12th, a panel of experts discussed the issue. The event began with remarks from Jonathan Mehta Stein, Executive Director of California Common Cause and a veteran civil rights attorney. He highlighted the severity of the situation: “We’re now entering the first AI election in which AI 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.” This captures the essence of the new challenges AI presents, including the potential for fake audio or videos—such as fabricated robocalls or misleading deep fakes—that could manipulate voter perceptions and undermine election integrity. While such phenomena are evident globally in places like Argentina and India, they are also occurring in the U.S., as seen with a deep fake of President Biden urging people not to vote in the Democratic primary earlier this year. Stein further elaborates on the localized impact of misinformation. “The dangers are less evident at the national level and more evident at the local level,” he notes. While national media may quickly debunk deep fakes involving high-profile figures, local political figures are more vulnerable. A deep fake of a city council member, for instance, could spread misinformation with minimal immediate scrutiny, damaging reputations and potentially influencing local elections. This issue is compounded by the proliferation of fake news websites, such as the Russia-linked Miami Chronicle, and counterfeit county election sites designed to mislead voters. Disinformation is particularly severe in other languages. Jinxia Niu, Program Manager of Chinese Digital Engagement at Chinese for Affirmative Action, highlighted the operational difficulties faced by fact-checkers working with content in Mandarin. “Currently, no one is managing the garbage cleaning work when it comes to language,” she explains. With only a handful of fact-checkers and a surge in AI-generated content, distinguishing genuine information from fake becomes increasingly challenging. Niu also points out the specific risks to immigrant communities, where AI scams and misinformation are prevalent. The difficulty in monitoring encrypted messaging apps like Telegram and WhatsApp exacerbates the problem, making it hard to combat misinformation within tightly knit communities. This concern is echoed by Brandon Silverman, a policy expert and former CEO of CrowdTangle. He emphasized the disproportionate impact on ethnic and marginalized communities. “One major issue is the disproportionate impact of misinformation on communities of color and ethnic communities,” he says. Silverman advocates for increased transparency and support for ethnic media to address these challenges. He notes that platforms often struggle with the “gray area” of misleading content that doesn’t strictly violate their rules but still confuses and misleads the public. Silverman also suggests practical solutions, such as directing digital advertising funds to support local journalism and ethnic media through taxes. This approach aims to bolster resources for communities most affected by misinformation. Additionally, he emphasizes the importance of scrutinizing influential narratives and engaging with consistent themes in AI-generated content rather than reacting to isolated instances of misinformation. As this issue continues to grow, it is imperative to remain skeptical, fact-check, and ensure information is verified before acceptance. |