The TikTok Marketing Podcast · Episode 100 · July 8, 2026 · 11 min
How Creators Are Selling Their AI Voice Clones
Episode 100 of the TikTok Marketing Podcast explores a growing creator economy trend: licensing AI-generated voice clones. Lucas and Luna break down the technology behind voice replication, the legal and ethical questions around consent and compensation, and how creators are turning their vocal identity into a scalable asset. They discuss real examples like the SAG-AFTRA deal with Replica Studios, the rise of synthetic voice platforms like ElevenLabs and Respeecher, and what this means for brand partnerships and content longevity. The hosts also consider the risks of deepfake misuse and creator burnout from licensing their voices. Tune in for a nuanced look at this frontier that blends marketing, tech, and intellectual property. #CreatorEconomy #AIVoice #VoiceCloning #ElevenLabs #Respeecher #SAGAFTRA #ReplicaStudios #SyntheticMedia #Licensing #IntellectualProperty #Deepfake #ContentCreation #TikTokMarketing #MarketingTrends #Business #FexingoBusiness #BusinessPodcast #Episode100 Keep every episode free: buymeacoffee.com/fexingo
Transcript
Lucas: So, there's this creator I've been following — she does cooking tutorials on TikTok, really distinctive voice, warm and kind of gravelly. She recently announced she's licensing an ai generated version of her voice to a meal kit company, so the brand's automated phone system and video ads now literally sound like her. Luna: Wait, her actual voice? Like, cloned? Lucas: Exactly. She recorded about three hours of scripted audio, the company fed it into a voice synthesis model, and now they can generate new lines in her voice without ever bringing her back into a studio. She gets a flat licensing fee plus royalties per month. And she's not alone — this is becoming a real revenue stream for creators. Luna: I've seen a few big creators mention it, but I didn't realize it was already commercialized. That's wild. Lucas: Yeah, and it ties directly into something we've talked about before on this show — creators turning their personal assets into scalable products. Voice is maybe the most personal asset of all. Before we go deeper, I want to say something about this episode itself. We're at number 100, which is kind of a milestone, and we've never run a single ad. That's intentional — we want these conversations to be exactly what we think is useful, not what a sponsor wants. If that approach resonates, the best way to support it is a one-time thing at buy me a coffee dot com slash fexingo. That link is the only thing we'll ever ask for. No tiers, no perks, just keeping this ad-free. Luna: Yeah, I love that we've kept it clean. And honestly, knowing listeners choose to support that — it's meaningful. Lucas: Alright, back to voice clones. The technology behind this has gotten shockingly good in the last two years. Companies like ElevenLabs and Respeecher can replicate a voice with as little as one minute of source audio. But for commercial licensing, most require several hours to get the intonation and emotion right. Luna: That raises an immediate question: once my voice is out there, who owns it? And can I stop someone from using it? Lucas: Right, and that's the frontier the industry is still figuring out. A few months ago, sag aftra — the actors' union — struck a deal with a company called Replica Studios. The agreement sets minimum compensation for voice actors who license their voice for video game and interactive media use. It's structured like a residual system: actors get paid each time their voice is used in a new project. Luna: So that's the union stepping in to create a framework. But most TikTok creators aren't unionized. They're negotiating these deals individually, which means the power balance is totally off. Lucas: Exactly. A creator with 5 million followers can command a decent fee — we're seeing numbers from five thousand to fifty thousand dollars for an annual license, plus royalties. But a mid-tier creator with 100,000 followers might get offered a flat five hundred bucks and zero ongoing payment. Luna: And the brand can then use that voice in perpetuity? That seems like a terrible deal for the creator. Lucas: It is if there's no sunset clause. The smarter contracts I've seen include a term limit — two years, for example — after which the rights revert to the creator. Some also restrict the contexts where the voice can be used: no political ads, no adult content, that kind of thing. Luna: What about the ethical side? I mean, there's already been cases of people using deepfakes to impersonate creators without permission. How do you prevent that when your voice is being actively cloned? Lucas: That's the biggest risk, and it's why some creators are hesitant. The platforms have started adding watermarking — ElevenLabs embeds an inaudible audio fingerprint in every generated clip so it can be traced back to the original model. But it's still early. If someone really wants to misuse a voice, they can scrape five minutes of public video and train their own model. Luna: So the solution has to be legal, not just technical. Creators need copyright protection for their voice itself. Lucas: That's an open legal question. The U.S. Copyright Office has ruled that ai generated works aren't copyrightable, but the underlying human performance might be. Some states have right-of-publicity laws that cover voice, but they vary wildly. Tennessee just passed the ELVIS Act, which explicitly protects voice as a property right. Luna: ELVIS — that's 'Ensuring Likeness Voice and Image Security,' right? It's aimed at preventing unauthorized AI impersonations. Lucas: Exactly. And it's the first state law that specifically calls out synthetic voice cloning as a violation of personality rights. I think we'll see more states follow, and eventually federal legislation. But for now, creators are largely on their own. Luna: So what's the smart play for a creator who wants to explore this? How do they even start? Lucas: First, they should never give away raw audio without a contract. Record a demo specifically for the licensing deal — don't hand over your entire TikTok archive. Second, the contract should specify the use case, territory, and duration. Third, include a clause that requires the licensee to implement technical protections, like watermarking. Luna: And what about exclusivity? If I license my voice to a meal kit company, can I also license it to a podcast app? Lucas: That depends on negotiation. Some deals are non-exclusive, which is better for the creator because they can stack multiple income streams. Others offer a premium for exclusivity. The creator I mentioned at the top — she opted for non-exclusive, and she's now in talks with three other brands. Luna: I want to zoom out a bit. This feels like a natural evolution of the creator economy. First it was photos, then videos, then data, then writing styles. Now it's the actual sound of a human voice. What does that mean for authenticity? I mean, part of a creator's appeal is that it's them, live, unfiltered. Lucas: It's a trade-off. On one hand, a voice clone lets a creator be 'present' in multiple places at once — a brand ad, an audiobook, a video game — without spending hours in a studio. That's scalability. On the other hand, if every brand interaction sounds like the creator but isn't actually them, does that dilute the personal connection? Luna: I guess it depends on disclosure. If brands clearly label 'this is an ai generated voice,' then the audience can decide. But if it's hidden, it feels deceptive. Lucas: And that's where regulation might come in. The FTC has already warned about ai generated endorsements — they require clear disclosure if the voice isn't the actual person. So honest labeling is likely to become mandatory. Luna: Let's talk about another angle: longevity. A creator's voice could outlive them, right? Like, if they license it now, the brand could keep using it even after the creator stops making content — or worse, after they die. Lucas: That's a real consideration, and it's already happening in music. There's a whole debate around posthumous voice cloning — think of the hologram Tupac at Coachella, but now it's audio-only and far more lifelike. Some creators are including a clause that terminates the license upon their death, or that requires estate approval. Luna: So it's not just a contract issue — it's an estate planning issue. That's heavy for a 22-year-old making dance videos. Lucas: It is. But the smart ones are thinking about it. I talked to a creator who has a clause that any royalties after her death go to a charity. She said it makes her feel like her voice can keep doing good even when she's gone. Luna: That's actually beautiful. Let's pivot to the brand side. Why would a brand want a creator's voice instead of just hiring a voice actor? Lucas: Two reasons. One: fan loyalty. If you're a meal kit brand and your target audience loves this cooking creator, hearing her voice in your ads builds instant trust. Two: consistency. The creator's voice becomes a sonic brand asset — like Intel's bong or Netflix's 'ta-dum' — but with personality. Luna: So it's not just about the voice itself, it's about the association. That makes sense. But how do brands measure the ROI of a voice clone? It's not like a click-through rate. Lucas: They're using brand lift studies, same as with any influencer campaign. They survey audiences before and after exposure to see if recall and favorability increase. Some are also A/B testing — same ad script, one with the creator's voice, one with a generic voice actor. Early results show the creator's voice can lift purchase intent by 15 to 20 percent. Luna: That's significant. So there's a real business case. But I wonder if we're heading toward a world where every brand voice is a cloned creator voice, and actual human voice actors get squeezed out. Lucas: It's a valid concern. Voice acting is already a tough profession. But I think there will always be demand for original human performances — for animation, for complex emotional scenes, for live events. The clone works best for predictable, repetitive use cases: phone systems, video narrations, training modules. Luna: So it's more like a replacement for text to speech robots, not for actors. That's a more optimistic framing. Lucas: Exactly. And creators themselves are starting to build businesses around this. I know of a TikToker who set up a marketplace where other creators can list their voices for licensing — she takes a 20 percent cut. It's like an app store for voices. Luna: That's clever. So she's not just monetizing her own voice, she's creating a platform. That's the kind of entrepreneurial thinking we love on this show. Lucas: For sure. And it points to where we're heading: voice as an asset class. In the next few years, I bet we'll see voice licensing become as standard as image licensing. Creators will have a portfolio — their photo rights, their video rights, their data rights, and their voice rights. Luna: It's a lot to manage. But also a lot of opportunity. So for episode 100, I think we've hit on something genuinely new — not just a trend, but a shift in how creators think about their own identity. Lucas: Yeah. It's one thing to sell a video or a template. It's another to sell the thing that makes you unmistakably you. And the fact that creators are navigating this with contracts, ethics, and even estate planning — it shows how mature this economy has become. Luna: Definitely. And I'm curious to see how the legal landscape catches up. I'd bet by this time next year, we'll have a federal framework or at least a model law. Lucas: I hope so. Because if we get it right, creators can benefit enormously. If we get it wrong, we could see a wave of exploitation. The next hundred episodes will tell. Luna: Alright, that wraps it up for episode 100. Thanks for listening, and as always, we'll be back tomorrow with another angle. Lucas: See you then.