×

Customers

[email protected]

UK: +44 (0)203 603 3676

Talent

[email protected]

UK: +44 (0)203 603 3676

Industry

Rise of the Machines: What is Voice Cloning?


By Michael Sum | 7th October 2021

In an age of self-driving cars and facial recognition on smartphones, how do you know the voice you are listening to is even real?

A concept that was once considered a hallmark of science-fiction has become eerily relevant in recent years: the act of cloning. (Dramatic dun-dun-duuuun).

Okay, so not exactly "cloning" in any genetic sense per se; less Blade Runner, more The Matrix. Voice cloning refers to the artificial replication of someone's speech to be employed to say anything as if the person had said it themselves. It's enough to make you question the audio you hear online.

Sounds fun, right? Getting a celebrity or politician to say anything! But think a bit more about the end of that sentence; say anything

Is the ownership of our own voices at risk?

That is why we are here today, to investigate the issue and discover what voice cloning is and how you can protect yourself against it.


What is Voice Cloning?

It's already here. Voice cloning services exist across the internet, many of which are free to use.

Voice cloning is part of the Deepfake trend that has swept the web. Deepfakes are a special kind of disinformation that uses artificial intelligence to seamlessly replicate a person's face and transplant it onto someone else—applications of this span from harmless fun to widespread abuse of someone's agency over their own body.

Yep, inappropriate and unauthorised usage of a person's face is a genuine threat. Deepfakes are responsible for unparalleled risks to security, ownership, and media authenticity – all traits that apply to the definition of voice cloning.

Instances of voice cloning are remarkable and are unsettling in how commonplace they can be. AI and Text-to-Speech voice-over can be valuable tools to generate voice-over procedurally. Applications include testing creative copy or quickly and cheaply making written content accessible to the visually impaired. 

However, this presumes that original voice talent is satisfied with the arrangement and is fairly renumerated. With advanced voice cloning techniques, this may become difficult to control.

So yeah, we have a problem. As technology progresses and advanced functions become more accessible for the public. Anyone with sufficient content online is at risk of voice cloning.  Celebrities and politicians are the obvious targets, but there is plenty of voice actors at risk too.


Are there genuine uses for Voice Cloning?

Let's quickly acknowledge the possibilities that come with voice cloning. The potential applications for a synthetic and adaptable voice are limitless – it speaks to the core of the voice-over industry itself. Not all of these are necessarily harmful either.

For instance, modern role-playing video games (RPGs) have geared their protagonists away from traditionally silent characters and given them voices. These characters can have hundreds of thousands of lines to record.

While voiced characters provide emotional depth, they also limit the player's options. A top-tier synthetic voice could circumvent the need to cut corners and provides the best of both worlds. But it's unlikely that AI voices will be able to match the emotional depth or range of a human voice anytime soon. 

So, what else?

Voice cloning is more likely to have used in both pre and post-production. Imagine being able to hear voices read your script without the time-consuming task of requesting auditions? Or using a clone of the actors' voices to test out script rewrites. In post-production, the ability to get pickups recorded without scheduling studio time could also be a boon. And of course, in the rare case, an actor dies mid-production, a clone of their voice could help fill out any missing audio.


The Dangers of Voice Cloning

The future of the voice-over industry may be at the mercy of this new technology. But what's the long-term harm?

What is voice cloning's possible existential threat to the ability of voice actors to work safely? Let's find out.

The Diminished Worth of a Voice

The most apparent threat to the voice-over industry is plain. Why would a company go out of their way to hire a voice actor for a project when they could get a synthesised voice for a much lower cost.

As I said above, there is no way that an automated voice could match the emotional range of a human being - for now and probably the foreseeable future. However, technological advancement is a consummate trickster and makes the unexpected a reality. 

Although, minor roles in animations, commercials, and video games may use voice cloning for their less heard parts - sidestepping the casting and recording process entirely. A scary notion for us in the industry! 



The Fanmade Quest Mod for The Witcher 3 uses voice cloning to replicate protagonist Geralt of Rivia's signature speech.

Some fan modifications for existing games add in new content. The most impressive has started implementing services like CyberVoice to clone the in-game characters to enhance the immersion.


The Reputation of a Voice Actor

Voice actors live and die by their reputation. It can take years to break into any acting career, defined by hard work and training. As competition continues to rise in voice-over industries, your reputation is all the more important! 

Let's then imagine your voice is utilised to create a sound clip of you saying something horrendous. Something you wouldn't want your mum or your kids to hear. If anyone believes it is you, it could have disastrous consequences for your career.

The internet is full of people with bad intentions. And they may take great pleasure in using a voice clone to make it appear you said something explicit or offensive. 

This is a growing trend across all Deepfakes, placing people in nefarious predicaments. Protecting oneself against voice cloning and Deepfakes will be an ever-increasing societal problem in the coming years. I look forward to witnessing future tech entrepreneurs combat the threat!


Legal Ramifications and Ownership

It's tricky to discuss ownership in this context. The ownership limits are clear in media content and performances - you perform once you have agreed to a fair price to both parties! 

Voice cloning typically will circumvent any negotiations with voice actors and implement the automated voice without the original owner's consent, which is troubling, especially where costs are concerned.

Bev Standing, a professional voice-over artist, was driven to suing social media goliath TikTok after using her voice in their text-to-speech functionality without proper compensation. While she was paid for the initial recording session, her voice was repurposed for usage across millions of videos without payment included in her fees.

Similarly, the original voice behind Apple's virtual assistant, Siri, was not paid for being Siri – she received payment for her initial recording session. Still, these payments did not include adequate remuneration for her voice playing on countless iPhones.



The voice of Siri was not made aware of the role her voice would play.


The Right to Say No

Everyone deserves the right to say no. The right to deny a request or reject a project should always be an option for any voice. Perhaps you have an ideological concern with a promo for a specific political party? Or maybe an organisation seems dubious, and you would instead protect your reputation from affiliation with them. 

Whatever the reason, this choice is your right. And when voice cloning means that anyone can access a voice and use it for any nefarious purpose - without consent - it is a scary world indeed. 


How can you protect your voice?

Protecting your voice from being exploited is the real trick. Some may even want to get ahead, cloning their own voice. But the most surefire way to protect a voice talent against unethical replication and misuse is to promote legislation to protect against Deepfakes and methods included in voice cloning. 

Legal precedent derives from cases that act as references for future disputes. Like Assembly Bill 602 in California, which protects Deepfake victims from pornographic exploitation - there is space to define this legal grey area with vigilance. There are constantly cases like this, popping up as technology continues to press on. 

The issue of voice cloning is often compared with the music industry of the 20th century. As music became recordable in portable forms, the need for a live band in a bar or music player accompanying a film diminished, and musicians had to adapt. 

While there are similarities, what is a more appropriate comparison is a situation where the essence of a band could be replicated by a machine and songs procedurally generated to mimic their music perfectly, which is a problem that is harder to solve.

Protecting your vocal brand is essential - so you should always stay up to date on the developments in voice cloning tech! 

One easy way to protect yourself is to limit the places your voice can be heard in isolation. Ensure your demos have background music and effects, and this will make it more difficult to clone your voice (though not impossible).

I hope this helped answer the question 'what is voice cloning?'. Now, why not listen to some amazing sounding human voice actors.


Authentic Human Professional Voice Actors


Sometimes we include links to online retail stores such as Amazon. As an Amazon Associate, if you click on a link and make a purchase, we may receive a small commission at no additional cost to you.


Michael Sum

Michael Sum

Marketing Specialist at Voquent.

About Author

Michael Sum

Michael Sum

Marketing Specialist at Voquent.






Related Articles