In June, SiriusXM Media announced it has teamed with software developer Narrativ to deliver AI-generated voice replicas to advertisers.
With interest in the uses of AI in the radio industry so strong, Radio World wanted to learn more, so we reached out to Lizzie Collins and Matt Mohun of SiriusXM Media.
SiriusXM Media has an in-house creative agency, Studio Resonate, that hosts a self-service creative tool called AdMaker. The new partnership with Narrativ will allow SiriusXM Media’s clients to choose AI voice replicas in AdMaker when requesting audio ad creative.
Clients will be able to choose an AI voice replica from a roster of Studio Resonate voice actors who opt in to Narrativ. The staff of Studio Resonate will oversee the creative production. Each ad must be approved by the voice actor and advertiser before it goes live.
Actors who opt in have the opportunity to earn income through this licensing of their voice replicas. Narrativ signed a deal with SAG-AFTRA last year to protect and monetize union talent voice replicas.
Our Q&A follows. Lizzie Collins is SiriusXM Media’s senior vice president, B2B marketing and ad innovation. Matt Mohun is VP, Creative Services: Studio Resonate at SiriusXM Media.
Radio World: How does adding AI-generated voice replicas help SiriusXM Media advance its business goals?
Lizzie Collins: AI can be an extremely powerful tool — when used correctly. Our focus is on using AI to enhance our experiences that are led by human touch, and to find the white space in the audio advertising industry where it can be truly additive.
At SiriusXM Media — which is the sales organization for SiriusXM, Pandora and extensive streaming and podcast networks — we want to make the ad creative process easier, more cost-effective and more flexible for advertisers, which allows us to welcome more brands into the audio space as we look to close the gap between time spent with digital audio (25%) and ad spend with the medium (8%).
Through our partnership with Narrativ to bring AI voice replicas to audio ads, we can meet growing advertiser demands for more personalized, professionally produced audio creative at scale — something particularly challenging for smaller brands due to traditionally high costs, complexity and limited flexibility in production processes.
Our work with Narrativ addresses these barriers, providing a solution that is affordable, flexible and respectful of voice talent’s rights and revenue opportunities, ultimately making audio advertising creative much more accessible to brands of all sizes and helping us grow audio as a whole.
RW: Has the audio advertising industry in general gotten very far in adopting AI-based tools? In what form is AI changing this business sector?
Matt Mohun: AI is certainly taking off in audio, and it’s exciting to see the different ways companies are experimenting.
I will say that many AI audio tools seem focused primarily on automation alone, but our initiative strategically combines ethical AI implementation and tech, robust engagement with our voice actors, our in-house creative agency Studio Resonate’s human creative direction, and innovative advertising formats.
Narrativ is the first AI voice marketplace to be endorsed by SAG-AFTRA, ensuring that voice talent retains full agency over their replicas It’s incredibly important to us that we deliver AI-voiced ads that meet the same high creative and quality standards as those voiced by humans, while making sure there is consent, compensation and control when it comes to the use of replicas.
RW: Protecting voice talent is an important aspect of the arrangement.
Mohun: We chose to partner with Narrativ specifically because they share our belief in the power of the human voice. This deal allows us to continue harnessing that by finding ways to provide additional income to the great voice actors we work with every day.
Narrativ is all about doing AI voice replicas ethically, which means making sure there is consent, compensation and control when it comes to the use of replicas. Their deal with SAG-AFTRA also speaks to their high standards working with voice talent.
It’s important to note this is a 100% opt-in process for Studio Resonate’s roster of voice actors. If a voice actor decides to opt in, they can earn passive income through their voice replicas, and they maintain full control over voice usage, approving or denying every ad using their voice replica.
We work with over 200 voice actors, and many of them have opted in to have their voices replicated. We believe this partnership will allow voice actors who opt in to diversify their reach and scale their talent like never before, opening up new monetization opportunities.
RW: In what other areas of your business is the company deploying AI-based tools and services?
Collins: AI and ML have been at the center of our business for many years. In fact, Pandora has long been a pioneer in this space, with its algorithmically driven radio and the innovative Music Genome Project.
Today, across SiriusXM, we’re using AI in different ways to make it easier for people to find and interact with our business and our content, from personalization to customer agents to, in our advertising business, optimizing ad delivery to the right listener or bringing these AI voice replicas into ad creative. We’re committed to using AI to elevate every aspect of the listener experience, always guided by a human touch.
RW: What else would radio peers want to know about this initiative?
Collins: The opportunities for scale and flexibility are massive here, for both the voice actor and for the brand. Imagine a grocery store that wants to do hundreds, or even thousands, versions of an ad with a different call to action or a different location it wants to promote. Realistically, voice actors don’t want to spend eight hours a day for two weeks straight recording 4,000 location tags. This initiative allows us to make the re-editing process much more efficient for the voice actor, and much easier for the brand.
A new Radio World ebook explores how technologies based in AI are being deployed across radio (find it here).
