Progress has been made, but SAG-AFTRA says the two sides in the dispute are still far apart.
The SAG-AFTRA union says “progress has been made” in negotiations over the videogame voice and motion capture actors strike that was called in July 2024, but warned that the most recent proposal submitted by the bargaining group is “filled with alarming loopholes that will leave our members vulnerable to AI abuse.”
The strike, which officially began on July 26, 2024, calls for “fair compensation and the right of informed consent for the AI use of [actors’] faces, voices, and bodies,” SAG-AFTRA national executive director Duncan Crabtree-Ireland said at the time. The game makers negotiating with the union said they’ve offered “meaningful AI protections that include requiring consent and fair compensation,” but the union feels otherwise, claiming game companies involved in the negotiations “refuse to plainly affirm, in clear and enforceable language, that they will protect all performers covered by this contract in their AI language.”
That clearly continues to be the case. SAG-AFTRA said in a new statement that the bargaining group representing game companies “would have you believe that we are close to reaching a deal,” but insisted that’s not actually the case. It also shared a chart of proposals to demonstrate “how far apart we remain on fundamental A.I. protections for all performers.”
“They want to use all past performances and any performance they can source from outside the contract without any of the protections being bargained at all,” SAG-AFTRA said.
“You could be told nothing about your replica being used, offered nothing in the way of payment, and you could do nothing about it. They want to be able to make your replica continue to work, as you, during a future strike, whether you like it or not. And once you’ve given your specific consent for how your replica can be used, they refuse to tell you what they actually did with it.”
The rise of AI is a major concern for voice and mocap actors in the game business, and rightly so. If game developers can reduce their budgets by using machine-generated voices rather than those of real people, it’s a safe bet that they will. Whether generative AI can match a genuine human performance is a matter of debate, but there’s no question that the technology is getting better, and sooner or later the difference may become indistinguishable—or if it can be distinguished, a significant portion of the gaming audience simply may not care.
We’re not quite at that point yet, as the strong negative reaction to Sony’s recently leaked work on an AI-powered Aloy makes clear. But that experimentation also makes clear that game makers are pursuing this sort of work, and while the very unnatural voice of that chatbot can be easily handwaved as a placeholder, it’s work toward a goal, and those systems will get better.
SAG-AFTRA warned that game companies are hoping union members “will turn on each other” as the strike drags out, but the game makers themselves are facing growing pressure: League of Legends studio Riot Games was recently forced to recycle old voiceovers for some of its English-language skins due to the strike, and more recently Bungie said Destiny 2’s Heresy episode launched with some voice lines missing.
In February, more than 30 members of Apex Legends’ French cast—not members of SAG-AFTRA, but you know France and its unions—reportedly refused to sign a new contract because it required them to allow their work to be used to train generative AI systems.
“With their previously signed projects dragging their way through the production pipeline, employers are feeling the squeeze from the strike, as SAG-AFTRA members who work in video games continue to stand together and refuse to work without adequate protections,” the union wrote. “This is causing employers to seek other performers they can exploit to fill those roles, including those who don’t typically perform in games.
“If you’re approached for such a role, we urge you to seriously consider the consequences. Not only would you be undermining the efforts of your fellow members, but you would be putting yourself at risk by working without protections against AI misuse. And ‘AI misuse’ is just a nice way of saying that these companies want to use your performance to replace you—without consent or compensation.”
About Post Author
