As a result of AI, Val Kilmer’s voice has been restored. Critics, on the other hand, are concerned that the technology could be abused.
Sonantic, a British firm, recreated the actor’s original voice by using old films of him.
‘My voice was gone from me. Val Kilmer’s voice was heard in a YouTube video last week saying, “People around me have a hard time understanding me while I’m talking.”
He lost his natural voice after a throat cancer surgery in 2015, which ruined his career and changed the way he communicates forever. A voice cloning company in the U.K. called Sonantic helped Kilmer to talk again late last year.
Siri and Alexa, two of the most popular artificial voices based on actual people, sound false. For digital assistants, video games and movie studios, a new wave of start-ups is developing artificially intelligent voice cloning services.
Using deepfakes, a technology that employs artificial intelligence to modify content to appear and sound authentic, the created voices have become more realistic. Sometimes it’s hard to detect the difference between human and synthetic voices in this media.

His reps contacted Sonantic five years after Kilmer’s operation to digitally resurrect his lost voice.
Zeena Qureshi, CEO and co-founder of Sonantic, said, “So that’s what we did.” In order to continue creating, Val’s team needed to give him his voice back.
On December 2020, Kilmer finished filming “Val,” a documentary about his Hollywood career and disease.
There was no mention of Sonantic’s AI technology in the film. There is however a YouTube clip that has more than 18,000 views.
An undisclosed voice-cloning program was utilized by documentary director Morgan Neville to imitate the late chef Anthony Bourdain in his commercial film “The Road Runner” last month, according to Neville. Ottavia Bourdain, the late actor’s widow, took issue with Neville’s claim that she was approached about re-creating her husband’s voice with AI.
As for the other actors Sonantic is working with, they declined to comment. Obsidian Entertainment and Remedy Games are among the gaming companies that the three-year-old startup works with. It licenses its synthetic voice service to studios, allowing them to edit and guide artificial voices in the same way that directors may influence human performers.
John Flynn, the company’s chief technical officer, compares it to Photoshop for voice.
A voice may usually be re-created within 24 hours given three hours of recordings. Sonantic had to re-create Kilmer’s voice using less than 30 minutes of audio because of movie licensing restrictions.
In order to eliminate background noise, engineers took samples from old film and “cleaned” them. Their computers learned to speak by listening to recordings, said Flynn. They constructed a screenplay based on the material, connected audio and text together in short pieces and ran the data through their “speech engine” algorithms.
An official statement says the voice engine can “illustrate severe rage and emotional suffering” by using the printed words as triggers. To demonstrate the service’s capabilities, Sonantic demonstrated a heated dispute in April. One voice talks to the other, but it quickly turns into a shouting match. Actors’ vocal cords would be preserved and they would be able to earn passive revenue, the business said in a press statement.
Firm says it developed 40 different versions of Kilmer’s voice and selected one that captured his expression the best. A text-to-speech algorithm that Sonatic claims can match Kilmer’s projection levels and emotion was the outcome.
As a result of the software, Kilmer’s former subtleties in speech, expression, and tone should be captured. Sonantic says Kilmer is free to use the technology however he wants.
„Only he has the right to use this model. She added, “He can utilize it for personal or business use.”
For persons who have difficulties speaking, or actors who need to rest their vocal cords after long screaming sessions in the studio, the technique may be useful.
There are legal, ethical and economic worries concerning the technology as well. This is especially true for voice performers who fear that their livelihood may be threatened.
Donald Trump and Barack Obama were filmed using deepfake technology, which highlights the perils of technology that makes it appear as if individuals are speaking things they never uttered.
According to voice actor Jay Britton of “Go! Go! Cory Carson” on Nexflix and many other video games, “I get to determine whether or not I support the content.” To inform a voice actor that their voice is out there speaking things that they might not necessarily support would be awful.
As stated by Sonantic, its product is not intended to replace actual performers. Website claims that device can “lower production timeframes from months to minutes,” promising “compelling, lifelike performances for games and films with fully expressive AI-generated voices,” premises that might reduce the number of hours human actors are paid to perform in studio.
Synthetic voices are not prohibited in the United States by law.
Although it is illegal for anybody to profit on the likeness of a famous person, there are legal safeguards in place. Tom Waits, a singer with a gruff voice, sued Dorito Lay in the 1990s for using a sound-alike voice in an ad and won $2.6 million in damages.
In New York City, intellectual property lawyer Peter Raymond warns that “if (businesses) reproduce voices of famous persons without authorization, they may be infringing the right to privacy and subjecting themselves to a lawsuit.” In the case of parody or creative routine, it’s not illegal. This is illegal if it’s for commercial gain.”
Kilmer praised Sonantic’s software, saying it “masterfully restored my voice in a way I’d never believed possible.”