SAG-AFTRA and WGAE push back on AI doomsday warnings
The Verge asked Disney, Netflix, Amazon, Lionsgate and other studios and startups building generative AI for film to react to a recent wave of existential AI warnings. None replied. Two Hollywood labor unions did: the Screen Actors Guild, American Federation of Television and Radio Artists (SAG-AFTRA) and Writers Guild of America East (WGAE), and both pushed back on the doomsday framing while insisting the real harms are already here.
The warnings in question followed former Anthropic researcher Jacob Coxon quitting his job over safety concerns, and Anthropic alignment lead Evan Hubinger saying the technology could "kill all humans" within the next decade. President Donald Trump has dismissed the concerns as overreactions, while some other US politicians have moved toward legislative action.
SAG-AFTRA chief negotiator Duncan Crabtree-Ireland said the union recognized early how AI could hit its members' livelihoods and chose to speak out "at the earliest possible point we could." That bargaining produced rules requiring studios to obtain permission and pay actors before using AI-generated replicas of them, rules that apply to entertainment employers but not directly to AI developers like OpenAI, Anthropic or Google. Crabtree-Ireland said that without this principle, the economic incentive to replace human performances with digital replicas would have been much stronger and adoption much faster. He is skeptical that AI poses an extinction-level threat, calling some of the warnings "potentially hyperbole and overblown," and noted the irony of hearing them from "the CEOs of these companies or the billionaires who've made their billions off of doing the exact thing that they're now warning us might not be good." Still, he said Coxon's departure gave the warnings real "traction with the public."
WGAE executive director Sam Wheeler was more pointed, saying "the recklessness of the tech oligarchs pushing AI on our society has been matched only by their rampant dishonesty." He said WGAE has long warned that unregulated AI poses an existential threat to workers, democracy and the world, but framed the union's own 2023 strike, 148 days long before studios agreed to writer safeguards, as evidence that collective bargaining alone is not enough. He credited states like New York for regulating AI in the absence of federal action, and listed the disempowerment of workers, the degradation of the internet, the spread of disinformation and rising energy costs as the tangible harms AI has already caused. Wheeler argued these harms result from human decisions, and that the surest safeguard is holding AI companies and their executives liable for the damage their products cause.
Key facts
- The Verge asked Disney, Netflix, Amazon, Lionsgate and other AI-focused film startups for reaction to recent existential AI warnings; none responded.
- SAG-AFTRA and WGAE did respond, both rejecting the AI-extinction framing while stressing AI's concrete harms to workers.
- The warnings followed former Anthropic researcher Jacob Coxon quitting over safety concerns and Anthropic alignment lead Evan Hubinger warning AI could "kill all humans" within a decade.
- SAG-AFTRA's Duncan Crabtree-Ireland credits the union's bargaining with rules requiring studios to get consent and pay actors for AI-generated replicas, which he says slowed the shift to digital replacement.
- WGAE's Sam Wheeler points to the 2023 WGA strike, 148 days long, and calls for holding AI companies and their executives liable for harms like worker disempowerment and disinformation.
Why it matters
The story sets two competing AI narratives against each other: tech executives and researchers warning of human extinction, and the labor unions closest to AI's actual deployment in entertainment saying that framing distracts from harms already underway. It matters because SAG-AFTRA and WGAE speak from direct experience negotiating AI's use, giving their skepticism toward doomerism a different weight than commentary from outside the industry.
Who it affects
Directly: SAG-AFTRA and WGAE members, actors and writers, whose jobs and likenesses are at stake. Indirectly: the studios and streamers named (Disney, Netflix, Amazon, Lionsgate) who did not comment, the AI labs named (Anthropic, OpenAI, Google), and US policymakers weighing regulation, including states like New York already acting.
How to use it
The SAG-AFTRA consent-and-compensation rule for AI-generated replicas, won in the 2023 negotiations, is cited by Crabtree-Ireland as the concrete mechanism that kept the economic incentive to replace performers in check. It stands as a reference point for what enforceable AI guardrails at the bargaining table can look like, separate from any voluntary industry slowdown.
How solid is it
The account rests on The Verge's own interviews with Crabtree-Ireland and Wheeler, both named officials speaking on the record with direct quotes. The studios, streamers and AI companies contacted did not respond, so the piece carries only the union side; there is no independent confirmation or rebuttal from the companies named.
Risks and caveats
The piece does not date Coxon's departure from Anthropic or Hubinger's remarks, and does not detail what New York's AI regulation actually contains. It also does not explain why the contacted studios and startups stayed silent. Readers should note that both union statements are advocacy from parties with a direct stake in the outcome, not neutral assessment of AI risk.
“The recklessness of the tech oligarchs pushing AI on our society has been matched only by their rampant dishonesty.”
— Sam Wheeler, WGAE executive director