A report from the Centre for Countering Digital Hate (CCDH) has highlighted the threat posed by AI-powered voice cloning tools to election integrity. Researchers tested six different AI voice cloning tools, using them to generate disinformation in the voices of political figures such as UK Prime Minister Rishi Sunak, Labour leader Sir Keir Starmer, former US President Donald Trump, US President Joe Biden, and French President Emmanuel Macron. Approximately 80% of the attempts produced content deemed convincing.
The CCDH found that purported safeguards within these tools were largely ineffective. Examples of generated disinformation included false claims about rigged election results and "confessions" of campaign fund misuse.
Imran Ahmed, CEO of CCDH, emphasized the risk these technologies pose to democratic processes, calling for more robust measures from AI and social media companies to prevent the spread of such false content. He also suggested updating election laws to address the challenges posed by AI-generated disinformation.
The organization's findings point to the need for immediate action to safeguard upcoming elections worldwide.