U.S. intelligence officials have issued a warning about the increasing use of artificial intelligence (AI) by foreign entities to influence American voters ahead of the November 2024 election. Russia, Iran, and China are reportedly employing AI tools to amplify their efforts in swaying public opinion.
Russia, identified as the most adept and aggressive of the three, is particularly focused on discrediting Vice President Kamala Harris. Officials from the Office of the Director of National Intelligence (ODNI) and the FBI revealed that Moscow is emphasizing content that undermines the Democratic presidential candidate's personal qualities and policy positions.
The use of AI in these influence campaigns marks a significant development in the field of information warfare. Since the concept of AI was first introduced at a conference at Dartmouth College in 1956, its potential applications have expanded dramatically. The term "artificial intelligence" itself, coined by John McCarthy in 1955, has come to encompass a wide range of technologies and capabilities.
Russian operatives have reportedly doctored clips of Harris's speeches, altering her words using AI technology. They have also generated false text, photos, video, and audio content. One notable example is a viral staged video, attributed to Russian actors by Microsoft researchers, which falsely claimed that Harris had injured someone in a hit-and-run accident.
While AI is enhancing the capabilities of foreign influence operations, intelligence officials emphasize that it is more of an accelerant than a revolutionary change. For AI to have a more significant impact, adversaries would need to overcome usage restrictions on large language models, create their own models, or develop effective distribution methods in the target country.
The use of AI in influence campaigns is not limited to Russia. Iran has been leveraging AI to create fake news articles in English and Spanish, focusing on issues such as the conflict in Gaza. China, on the other hand, has been using AI to amplify content related to drug use, immigration, and abortion.
"Generative AI is an accelerant for influence efforts rather than a revolutionary change in them."
The involvement of AI in these campaigns reflects the rapid advancement of the technology. In 2020, OpenAI's GPT-3, trained on 175 billion parameters, demonstrated unprecedented language generation capabilities. The following year, an AI system wrote an entire article for The Guardian newspaper, showcasing the potential for AI to produce convincing human-like text.
U.S. intelligence officials are collaborating with AI companies and social media platforms to address these threats. However, they emphasize that decisions about content moderation and account management are left to the companies themselves.
As AI continues to evolve, its impact on elections and public discourse is likely to grow. The first AI Ethics guidelines, published by the European Commission in 2019, highlight the increasing awareness of the need for responsible AI development and use. As the November election approaches, vigilance and public awareness will be crucial in countering AI-enhanced foreign influence attempts.