What is voice cloning and how can it be used?

Voice cloning is actually self-descriptive. It’s the act of incorporating technology to recreate or clone a person’s voice.

Just a paragraph spoken by an individual is enough for voice cloning technology to not just match it in pitch, timbre and dialect, but to be programmed and adapted into wholly synthetic speech. 

Speech that can also be artificially injected with the required emotion, and which can be speeded up or slowed down.

This technology represents huge shortcuts and cost savings within the media industry, particularly relating to voiceover projects and animated films. Perhaps not good news for the voiceover actors, though, who may not be needed for any more than a film’s opening scene.

Created initially with the intent to help people with speech issues, such as aphasia or a loss of voice after illness/surgery, the technology is also being used to digitally disrupt certain sectors, in a similar way to how our music scene and book industry were disrupted by technology. A move that’s great for big corporations and conglomerates, but not so great for creators/artistes.

However, there’s another use for voice technology that’s even more concerning than the lost revenue of a handful of individuals holding Equity cards.

Crime.

Data has long been used as a weapon and it appears technology can be similarly manipulated.

There are many devices in existence that use an individual’s speech patterns to unlock their contents. Even phone calls could be synthetically generated…it’s not difficult to imagine someone parting with their life’s savings if they received a plea from a loved one in a simulated crisis. If you heard the distinct, unique voice of your parent/sibling/partner/friend on the phone—with clear distress in their voice, asking for financial assistance—wouldn’t you comply? It’s easy to discount an attempt to defraud you from some Nigerian prince you’ve never met…if you genuinely believed you knew the person on the other end of the phone, it would likely be a different matter. Your bank account could be emptied in minutes using this voice cloning method or you could unwittingly impart confidential information.

‘Deepfake audio’, as it’s sometimes described, could be used in blackmail cases, it could be manipulated to cause damage to a person’s reputation…it could even be the instigator of World War Three, if the voice of the President of the United States was simulated, giving the order to ‘push the big red button’. A scary thought indeed.

Just as technology looks to be wielding its power for the benefit of criminals, however, there are programs out there that are being used to combat such nefarious attempts. These work to detect irregularities, repetition or clues in the background noise that could suggest the voice in question is being synthetically produced.

The examples I’ve used are also exaggerated, if only to show the possibilities of such technology. There are various layers of authentication in most secure systems, with voice recognition/command being just one. Because, now that you know what voice cloning is, I don’t want you to be up all night. Sleep soundly, the robots are not taking over.