If you are battling a video game goblin who speaks with a Cockney accent, or asking a gruff Scottish blacksmith to forge a virtual sword, you might be hearing the voice of actor Andy Magee.
Except it's not quite Magee's voice. It's a synthetic voice clone generated by artificial intelligence.
Recommended Videos
As video game worlds get more expansive, some game studios are experimenting with AI tools to give voice to a potentially unlimited number of characters and conversations. It also saves time and money on the “vocal scratch” recordings game developers use as placeholders to test scenes and scripts.
The response from professional actors has been mixed. Some fear that AI voices could replace all but the most famous human actors if big studios have their way. Others, like Magee, have been willing to give it a try if they're fairly compensated and their voices aren't misused.
“I hadn’t really anticipated AI voices to be my break into the industry, but, alas, I was offered paid voice work, and I was grateful for any experience I could get at the time,” said Magee, who grew up in Northern Ireland and has previously worked as a craft brewery manager, delivery driver and farmer.
He now specializes in voicing a diverse range of characters from the British Isles, turning what he used to consider a party trick into a rewarding career.
AI voice clones don't have the best reputation, in part because they've been misused to create convincing deepfakes of real people — from U.S. President Joe Biden to the late Anthony Bourdain — saying things they never said. Some early attempts by independent developers to add them to video games have also been poorly received, both by gamers and actors — not all of whom consented to having their voices used in that way.
Most of the big studios haven't yet employed AI voices in a noticeable way and are still in ongoing negotiations on how to use them with Hollywood’s actors union, which also represents game performers. Concerns about how movie studios will use AI helped fuel last year's strikes by the Screen Actors Guild-American Federation of Television and Radio Artists but when it comes to game studios, the union is showing signs that a deal is likely.
Sarah Elmaleh, who has played the Cube Queen in Fortnite and numerous other high-profile roles in blockbuster and indie games, said she has “always been one of the more conservative voices” on AI-generated voices but now considers herself more agnostic.
“We’ve seen some uses where the (game developer's) interest was a shortcut that was exploitative and was not done in consultation with the actor,” said Elmaleh, who chairs SAG-AFTRA's negotiating committee for interactive media.
But in other cases, she said, the role of an AI voice is often invisible and used to clean up a recording in the later stages of production, or to make a character sound older or younger at a different stage of their virtual life.
“There are use cases that I would consider with the right developer, or that I simply feel that the developer should have the right to offer to an actor, and then an actor should have the right to consider that it can be done safely and fairly without exploiting them,” Elmaleh said.
SAG-AFTRA has already made a deal with one AI voice company, Replica Studios, announced last month at the CES gadget show in Las Vegas. The agreement — which SAG-AFTRA President Fran Drescher described as "a great example of AI being done right” — enables major studios to work with unionized actors to create and license a digital replica of their voice. It sets terms that also allow performers to opt out of having their voices used in perpetuity.
“Everyone says they’re doing it with ethics in mind,” but most are not and some are training their AI systems with voice data pulled off the internet without the speaker's permission, said Replica Studios CEO Shreyas Nivas.
Nivas said his company licenses characters for a period of time. To clone a voice, it will schedule a recording session and ask the actor to voice a script either in their regular voice or the voice of the character they are performing.
“They control whether they wish to go ahead with this,” he said. “It’s creating new revenue streams. We’re not replacing actors.”
It was Replica Studios that first reached out to Magee about a voice-over audio clip he had created demonstrating a Scottish accent. Working from his home studio in Vancouver, British Columbia, he's since created a number of AI replicas and pitched his own ideas for them. For each character he'll record lines with distinct emotions — some happy, some sad, some in battle duress. Each mood gets about 7,000 words, and the final audio dataset amounts to several hours covering all of a character's styles.
Once cloned, a paid subscriber of Replica's text-to-speech tool can make that voice say pretty much anything — within certain guidelines.
Magee said the experience has opened doors to a range of acting experiences that don't involve AI — including a role in the upcoming strategy game Godsworn.
Voice actor Zeke Alton, whose credits include more than a dozen roles in the Call of Duty military action franchise, hasn't yet agreed to lending his voice to an AI replica. But he understands why studios might want them as they try to scale up game franchises such as Baldur's Gate and Starfield where players can explore vast, open worlds and encounter elves, warlocks or aliens at every corner.
“How do you populate thousands of planets with walking, talking entities while paying every single actor for every single individual? That just becomes unreasonable at a point,” said Alton, who also sits on the SAG-AFTRA negotiating committee for interactive media.
Alton is also open to AI tools that reduce some of the most physically straining work in creating game characters — the grunts, shouts and other sounds of characters in battle, as well as the movements of jumping, striking, falling and dying required in motion-capture scenes.
“I’m one of those people that is not interested so much in banning AI,” Alton said. “I think there’s a way forward for the developers to get their tools and make their games better, while bringing along the performers so that we maintain the human artistry.”