February 9, 2024 at 03:58PM
Recent events highlight the increasing threat of AI-generated deepfakes targeting US elections. The use of modern generative AI platforms poses risks such as mass influence campaigns, automated trolling, and the spread of deepfake content. Experts stress the potential for AI to amplify phishing attacks and call for measures to defend against these evolving threats.
The meeting notes highlight significant concerns about the potential use of generative AI (GenAI) to disrupt the US elections, including the creation and dissemination of deepfake content. Malicious actors are increasingly leveraging advanced AI platforms to conduct mass influence campaigns, automated trolling, and the spread of disinformation.
These AI technologies pose substantial risks, with the potential for microtargeting individuals with highly convincing and tailored messaging, thereby influencing their beliefs and votes. The combination of social media and readily available deepfake technology could exacerbate polarization and lead to the creation of “bespoke realities,” where individuals believe in alternative facts.
Furthermore, the use of AI in crafting more believable, targeted phishing campaigns at scale amplifies concerns about election security. It is anticipated that cyber adversaries will increasingly adopt generative AI to make phishing and social engineering attacks more convincing, thereby infiltrating campaigns or impersonating candidates.
To defend against these threats, election officials and campaigns are advised to be aware of GenAI-powered risks and implement measures to safeguard against them. This includes controlling content from interviews and press conferences, training staff on AI-powered threats and enhancing cyber hygiene practices, as well as regulating and implementing protections against AI-generated deepfakes.
Additionally, recent legal developments, such as the FCC declaring AI-generated voice calls as “artificial” and illegal under the TCPA, provide new tools to combat fraudulent activities involving voice cloning technology.
Overall, the meeting notes emphasize the urgency of addressing the potential misuse of AI in election interference and the need for comprehensive strategies to mitigate these risks.