A long time ago, in a galaxy, not so far, the voice was the exclusive domain of talented humans. But today, in the era of the generative AI, even the emblematic voice of a legendary villain can be mentioned without a single vocal cord, raising major legal issues on the name, image and resemblance (null) of rights in the digital era.
Fortnite, the Epic Games video game juggernaut recently sparked a new non-playable character (NPC) who exploits power on the dark side using the vocal technology generated by Ai-Ai. In Fortnite, the voice of this antagonist casmed in black armor does not come from James Earl Jones, who expressed the character in the films, but from a machine formed to correspond to his tone, at his pace and his cadence. While Jones would have authorized the use of his voice, this decision triggered a storm of fire through the galaxy and among the lawyers on the future of the characters generated by the AI.
The basic problem is whether the voices recreated digitally, built by AI from past performance, are really protected by existing Nile frames. Under the current laws of the grip, individuals have the right to control the commercial use of their identity, including their voices. But when an in-depth learning model generates a new dialogue using a synthetic approximation of this voice, is it always “his voice”? Or has it become something new, a kind of legal clone soldier who escapes the reach of traditional protection?
The consent of the interpreter (or their field) is essential, but it is not a tip of Jedi that solves everything. Most zero laws have not been designed for this type of technology. Statutes, such as California Civil Code § 3344.1, protect deceased personalities, but only up to 70 years after death. Even then, the application can be more trembling than the goal of a stormtrooper. Worse, these laws do not always contemplate AI models that adapt and generate new content indefinitely, effectively transforming the resemblance of an interpreter into a Holocron of infinite expressions.
What if a license voice for a project ends up being used in a galaxy of games, films or even advertisements, without new compensation or surveillance? It’s not just speculative. Without clear contractual limits, a single vocal training set could lead to endless restarts of a character, long after the original artist or their succession would have said: “I find that your disturbing lack of faith.”
Developers and studios can see AI as a shortcut to hypero-spaces, time, money and coordination economy. But the use of AI to create derived vocal performance without the risks of transparent legal railings transforming zero rights into pantha fodder. It also opens the door to future disputes on the question of whether double digital requires the same license rigor as human artists, especially since the models of AI become capable of improvising and even emotional nuances.
This last episode NPC highlights a broader change in the entertainment industry, where the studios are turning more and more towards AI to preserve the legendary characters. But with great power, great responsibility (legal). If the digital dark side is not controlled, artists can lose control of their similarities, their voices and perhaps even their narrative heritage.
In short, the powers of AI are strong, but it must be balanced. Just as the Council has debated once if a young Padawan could trust with too much power, we must now decide whether the leisure of AI deserve to handle the resemblance of unlimited artists. Our helmet bad guy had an arc of buyout which finally restored order in the galaxy, but if the law does not follow the pace of technology, the entertainment industry may not be as lucky.
