Podcastle, a leader in AI voice cloning technology, has conducted a survey to identify which celebrities are perceived as most at risk of misuse through AI voice replication. The survey, involving 1,000 participants across the U.S., reveals notable insights into which celebrity voices are deemed easiest to clone and highlights the growing concerns surrounding this technology.
The survey found that nearly two-thirds of respondents (61%) believe Arnold Schwarzenegger’s voice is the most easily replicable, attributing this to his distinctive and instantly recognizable accent, which 86% of participants identified as unique. Additionally, 39% of respondents noted that Donald Trump's voice is also easy to imitate, thanks to its frequent exposure in the media. Kim Kardashian's voice was identified by 23% of participants as having a consistent tone and pitch, making it relatively simple to clone.
Actor Morgan Freeman emerged as the most favored choice for voice assistant technology, with 24% of respondents selecting him as their preferred voice. Freeman was also the top choice for a potential celebrity podcast host, with 11% of participants endorsing him, largely due to his reputation as a compelling storyteller who keeps audiences engaged.
The celebrities featured in the survey were chosen based on their combined search volume on Google and TikTok for “[celebrity name] + impression.” This selection criterion highlights the public's interest and concerns regarding the replication of celebrity voices.
The issue of unauthorized voice cloning has been highlighted recently when actress Scarlett Johansson expressed her shock and anger over a ChatGPT voice update that closely resembled hers. Johansson, who had previously declined an offer to voice ChatGPT, criticized the technology for its lack of consent and ethical considerations.
Voice cloning technology has sparked controversy, with some instances being banned. Earlier this year, ChatGPT introduced a text-to-voice feature that generated natural speech from brief audio clips but limited its availability to mitigate misuse.
Podcastle, which integrates generative AI features into its podcasting platform, including text-to-voice generation, raised $13.5M in a Series A funding round led by Mosaic Ventures earlier this year. The company acknowledges the need for improved safeguards and user education in the rapidly evolving field of AI voice cloning.
Arto Yeritsyan, CEO and founder of Podcastle, emphasized the company’s commitment to promoting AI voice technology responsibly. “Our goal is to balance innovation with responsibility, creating a platform where users can explore AI's benefits safely and creatively,” Yeritsyan stated.
