Key Points:
- With less than a year to the 2024 U.S. presidential election, there is increasing concern about the potential of accessible generative AI tools to disrupt the process and spread misinformation.
- Experts warn about domestic and foreign actors using AI to influence the election’s outcome and criticize social media platforms for letting down on trust and safety measures.
- Suggestions of using watermarking technology to authenticate legitimate images and videos from AI-generated deepfakes have been met with skepticism.
The 2024 United States Presidental Election is likely to face a significant threat from Artificial Intelligence (AI) driven disinformation, with growing concerns that accessible generative AI tools could severely disrupt the process. This potential threat was discussed openly at Fortune’s Brainstorm AI Conference in San Francisco, with several speakers expressing concern.
Common Sense Media’s Founder and CEO, Jim Steyer, predicts misinformation havoc during the elections. He warns of potential interference from both domestic and overseas entities swaying the election results. He critiqes social media platforms, such as “X” (formerly Twitter), and Facebook, for compromising trust and safety teams responsible for monitoring and controlling misinformation on their platforms.
Reid Hoffman, LinkedIn’s cofounder, expressed his concern over malicious actors exploiting AI to interfere with the upcoming election. He expressed skepticism towards suggestions of watermarking technology for segregating authentic images and videos from AI-created deepfakes. He emphasized the need for companies that command AI models, such as Google, OpenAI, and Microsoft, to establish a framework for this technology. However, he also noted that not all AI models implement such watermarking requirements.
Vinod Khosla, the influential Silicon Valley Investor and Sun Microsystems cofounder, estimates a 95% likelihood that generative AI will be a significant influence on the election. He predicts an engagement of over 100 million bots powered by persuasive AI interactively seeking to manipulate the election results. As these concerns continue to grow, stakeholders are awaiting effective solutions to control the AI-driven disinformation risk.