Deepfake technology has been a growing concern in recent years, with the use of AI to create realistic but false videos and images becoming more prevalent. The recent case of a deepfake of U.S. Rep. Rob Wittman being used in the Taiwan presidential elections to endorse a candidate has raised alarm bells for officials and advocates monitoring the spread of disinformation.
Experts like director of UC Berkeley’s Risk and Security Lab Leah Walker and Public Policy Student Fu-Yang Peng have been studying how AI and deepfakes are being used to influence elections around the world. With over 80 countries holding major elections this year, the need to understand and combat the spread of misinformation is more important than ever.
One example highlighted by Peng was a deepfake targeting Taiwan President Tsai Ing-Wen, spreading false information to damage her reputation and influence voter behavior. Walker emphasized the importance of studying smaller countries like Taiwan and Lithuania, as they often have better safeguards against disinformation and can provide valuable insights for larger elections like the upcoming U.S. presidential election.
UC Berkeley associate professor Andrew Reddie showcased the rapid advancements in AI deepfake technology in just 12 months, expressing concern about the potential consequences of these tools falling into the wrong hands. He stressed the need for AI companies to develop watermark features for AI imagery and video to help combat the spread of deepfakes.
While the deepfakes in Taiwan mostly originated from outside the country, watch groups were able to flag them and limit their impact. Peng and Walker noted the lack of robust watch groups in many countries, underscoring the importance of studying these cases to better prepare for the challenges that may arise in the lead-up to the November election.
As the technology behind deepfakes continues to evolve, experts are working diligently to stay ahead of the curve and protect the integrity of democratic processes worldwide. With the potential for foreign actors to exploit AI technology for malicious purposes, vigilance and proactive measures will be crucial in safeguarding elections from the influence of disinformation.