Why I Turned Down a $60,000 AI Voice Cloning Job

If you’re not a voice actor, you’ve probably never considered how AI voices can harm our industry; and there are hundreds of voice actors who refuse to touch AI voiceover work with a ten-foot-pole. I was one of those.

The general talk amongst voice actors several years ago was, “AI voices are bad, they can never replicate the emotional nuances that we, human voices can, and we should fight them at all costs.”

Now, we know that is unreasonable and unlikely and AI is happening and developing whether we like it or not. And some AI voices are getting pretty good and sounding more natural all the time.

So, I set out to educate myself on how to best embrace AI voice work and how to ethically sign on for some of these lucrative jobs.

For reference, there are hundreds of text-to-speech (TTS) and voice cloning jobs which can pay voice actors into the hundreds of thousands of dollars. There are AI voices for companies, cars, devices, chatbots and more. But you likely already know that.

What you may not know is the VO-community-agreed-upon-ethical-uses of AI voices: you should be paid fairly for the recording sessions, all usage including where and for how long your AI voice will be used should be clearly laid out and compensated and should you be tied to an exclusivity contract, you’d better demand the big bucks. There’s a lot more to it and if you want to know more, Google is your friend. Both the National Association of Voice Actors (NAVA) in the USA and the Canadian Association of Voice Actors (CAVA) have lots of literature about doing any kind of AI work. They also have a free AI rider, which you can include with your contract info that you send to clients.

Now to my own job: I’ve only actually applied to two or three AI voice cloning jobs (never text-to-speech) and that was only after combing through their job descriptions and usage details, along with the rate offered. This one was a maximum $60,000 for 200 hours of recording + in-perpetuity usage for an unnamed company, via a reputable pay-to-pay site. There weren't too many details. I thought, I’ll apply and see what the actual contract says. I was shortlisted and was sent the contract. There were red flags immediately:

1. The contract allowed the company to license or sell my voice to a third party: who monitors that company’s usage of my voice? I wouldn’t get a say in who they sell/give it to and I wouldn’t be paid for any of that usage.

2. There were strict behavioural restrictions: I wouldn’t be allowed to do any voice work for gambling companies (I already have an ongoing Lotto 649 campaign), anything sexual/erotic, offensive or anything involving hate speech. Of course, some of that is reasonable but who’s to say what is ‘offensive’? When language is that vague, it’s concerning because it's subjective. Plus, any restrictions could hinder my future career and, for $60,000 (or any amount, in my opinion), that shouldn’t be worth it to anyone.

3. The company can use my voice for whatever they like, for eternity: what if this harms my future career? What if it’s something I don’t want to represent? I’d have zero recourse if I signed this contract.

4. The company could terminate the contract at any time and yet still keep my voiceovers. How do I know they won’t still use them?

5. If the company breaches the contract, my only recourse is asking for more money (and there’s no guarantee I would get it). I’d have no option to terminate the contract from my end.

There were a lot more concerns and I flagged all of these to the P2P that I was working through. They wrote back and didn’t have answers to any of my questions or concerns because they are the ‘middle man’. So I immediately withdrew my application and warned the P2P that any voice actor should be seriously concerned about this contract.

With more of these jobs becoming available and AI voiceovers becoming more in demand, we have to accept that they are part of our industry now. But AI voices can’t be created without human voices (and hundreds of hours of recordings, at that!). I realize AI voices are attractive because it's cheaper and creates consistency for a brand, but it’s baffling to me that a P2P wouldn’t fight for more ethical contracts that protect the voice actor since they make their money off our backs every day. If we sell our voices so cheaply, and so often, to create AI voices, how many human-VO jobs are left for P2P’s to sell? And if companies are using AI more often, how does that make their brand look? At this moment in time, we can all tell what is AI and what isn’t although that difference is becoming harder to spot by the day. But right now, most people can agree that AI looks cheap and may cheapen the look of a brand.

Am I against AI? No, actually, I’m not. I truly believe it serves a purpose and there are some use cases where it’s a great option. But usage matters. Sure, having an AI chatbot is far cheaper for a company than having a team of live customer service reps, but the person helping to create that chatbot, or the person behind the AI voice needs to be paid fairly. If you’re planning to use it forever, you’d better plan on paying the voice actor forever, too.