In a recent study conducted by the Center for Countering Digital Hate, researchers found that publicly available artificial intelligence tools can easily be used to create convincing election lies in the voices of prominent political figures. The study tested six popular AI voice-cloning tools and found that they were able to generate realistic voice clones in 80% of cases. These tools were able to create fake statements about elections in the voices of political figures such as U.S. President Joe Biden and French President Emmanuel Macron.
The researchers highlighted a significant gap in safeguards against the misuse of AI-generated audio to deceive voters, raising concerns about the potential for widespread disinformation in the upcoming elections in the U.S. and the European Union. With limited regulations in place to prevent the abuse of these tools, voters are left vulnerable to AI-generated deception as technology continues to advance and become more accessible. The lack of self-regulation by companies providing these tools further exacerbates the risk of disinformation campaigns during important democratic processes.
The study revealed that some AI voice-cloning tools have minimal safety measures in place to prevent the cloning of politicians’ voices or the production of election disinformation. Despite some tools requiring users to upload unique audio samples as a safeguard, researchers found that these barriers could easily be circumvented using different AI voice-cloning tools. The ease with which these tools can create fake audio poses a significant threat to the integrity of democratic elections by spreading false information and manipulating public opinion.
While some companies, like ElevenLabs, have expressed a commitment to improving their safeguards and preventing the misuse of AI-generated audio, the overall lack of regulation and oversight in this space remains a concerning issue. Experts have warned that bad actors have already begun using AI-generated audio clips to influence elections, further highlighting the urgent need for tighter security measures and proactive transparency from AI voice-cloning platforms. Lawmakers are urged to take action to establish minimum standards for the use of AI in elections to protect against disinformation and safeguard the democratic process.
The potential impact of AI-generated disinformation on elections is significant, as it can erode public trust in the information they see and hear. As technology continues to advance rapidly, the threat posed by AI-generated media, including voice-cloning tools, requires immediate attention from regulators and tech industry leaders. The researchers stress the importance of implementing stricter regulations and security measures to prevent the spread of deceptive audio content that could undermine the democratic process and manipulate public opinion. The findings of this study underscore the urgent need for action to address the growing threat of AI-generated disinformation in elections around the world.
