Articles

Voice artists fear AI will ‘steal’ their voices

United Voice Artists (UVA), a global group of over 20 voice acting guilds, associations and unions, has begun a worldwide campaign to lobby governments to regulate the use of artificial intelligence (AI) in the creative industries.

In a statement, voice associations across the globe have raised concerns around the unregulated use of AI, highlighting the threat posed by AI-generated voices or cloned voices on their earnings.

Voice artists are concerned the increased use of unregulated voice-generating AI apps and platforms is capable of putting entire segments of the voice and art industries out of work.

The collaborative effort – called “Don’t steal our voices” − brings together prominent associations and unions from the European Union, including France, Spain, Italy, Germany, Austria, Belgium and Poland, as well as Switzerland, Turkey, US and South America.

“UVA calls on politicians and legislators to address the inherent risks, both legal and ethical, in the conception, training and marketing of AI-generated content,” says the UVA.

“Any use of AI technology to generate and clone human voices must be subject to the explicit consent of the voice-over artists and performers, which must therefore be in a position to refuse the use of their works and performances, past and future, for purposes not expressly authorised by them, and be offered practical solutions to ensure the effectiveness of this choice.”

Digital cloning is an emerging technology that involves deep-learning algorithms, used in the creation of a digital version of a person in the form of an image, avatar, video or audio, according to Wikipedia.

ITWeb recently reported on the opportunities and risks posed by AI’s ability to recreate and manipulate − in different ways − currently existing audio, photos and videos to create digital clones across many industries in SA.

Other concerns raised by UVA are those relating to the preservation of property rights and intellectual property rights. According to the body, currently, generative AI technology heavily depends on online sourcing to enhance its learning capabilities, which regularly involves the illicit scraping and use of copyrighted data and content.

“In terms of protecting personal data, voice-over artists alert data protection authorities that generative AI technologies are likely to feed off ‘sensitive’ personal data – voice is considered biometric data under the General Data Protection Regulation, in contexts and for purposes that are not yet well-defined nor transparent.”

More apps and platforms from the likes of artificial voice companies ElevenLabs and revoicer.com are offering a lower barrier of entry for those requiring celebrity voices, which are accessible at more affordable prices than using other human voices.

Discussing digital cloning, Flux Trends researcher and trend-spotter Faeeza Khan previously told ITWeb: “The advent of AI has brought with it numerous challenges for artists, celebrities, models and business people, who are now at risk of being impersonated in the digital realm, often without recompense.

“It’s already happening in South Africa and will continue to become problematic. We’ve already seen the case of the fake video, where president Cyril Ramaphosa spoke of a plan to tear down Voortrekker Monument and Loftus.”

Tags:
Comments
No Comments found
Leave a Comment