Hollywood’s entertainment unions are among the best-resourced, most organized labor groups in the country. In 2023, the WGA and SAG-AFTRA went on strike together for the first time in 63 years, largely over AI, and both came back with contract language most unions never get near. The WGA’s deal says AI can’t write or rewrite literary material, that AI-generated text doesn’t count as source material, writers get to decide whether AI touches their own work, and studios have to disclose when material contains it.
And yet, Hollywood reported losing over 6,700 jobs over the 12 months through May 2026, citing AI as part of the reason. Catie Maillard, Global VP of People at Data Society, is not surprised.
The WGA and SAG-AFTRA had just run a historic joint strike, and their own contract still couldn’t close the question that will matter most in five years: whether studios can train AI on writers’ past scripts, which the agreement leaves as “uncertain and rapidly developing.” If the two most organized entertainment unions in the country are still playing catch-up, most workplaces aren’t in the game at all.
“Unions have been trying to protect workers from AI, especially in Hollywood, over the last five years,” she says. “I think the unions have really been struggling to figure out how to react and how to get ahead of AI concerns, particularly while they’re still fighting the battle for living wages and what most countries consider standard protections. Hollywood has not been successful here – despite being one of the most powerful unions in the US, their leverage has waned over time. We’re finding that the powerful unions have at least some protections, though the majority of the United States that’s not unionized is a free for all. The United States government is not passing, and will likely not be passing, any regulations on businesses so many companies are looking more toward Europe, Australia, and Asia for responsible AI practices.”
Union membership in the US sat at 10.0% in 2025, propped up by the public sector, where 32.9% of workers are organized. Private-sector union membership, the sector most AI-affected industries sit in, was 5.9%. There is no federal law in the US governing how employers can use AI in hiring, monitoring, or workforce decisions. Absent a union contract, an employee’s only real protections are whatever the employer decides to offer.
That’s the contrast behind Maillard’s point about companies looking abroad. The EU AI Act classifies AI systems used in hiring, performance evaluation, and worker monitoring as “high-risk,” which triggers mandatory risk assessments, human oversight, and disclosure requirements before an employer can deploy them. Australia has been moving in the same direction on workplace surveillance and AI use. Neither is a US requirement. A US company that wants that kind of structure right now has to build it voluntarily.
The second half of Maillard’s point cuts the other way: even where AI does get blamed for job losses, it’s often not the direct cause.
“This has been a horrible year of mass layoffs in the US’,” she says. “We’ve also seen the headlines around large company layoffs often pointing to AI as the reason behind the decision, but in reality it’s often not as direct. It’s more likely, ‘we changed our business strategy, we’re offshoring, we want to direct money somewhere else.’ Though I haven’t often seen companies explicitly saying, ‘we found efficiencies and can lay off employees.’ That’s very few of the actual layoffs I’m seeing. A lot of times I’m seeing AI as either a smokescreen for other business strategy changes, or a tool that highlights imperfections in business strategy that then need to be challenged and changed.”
Researchers watching the same layoffs from the outside are describing the same pattern. Evan Sohn of Revelio Labs put it plainly: payroll is being converted into capital expenditure. Companies are cutting headcount while ramping up spending on AI infrastructure and data centers. Georgetown’s Jason Schloetzer argues the AI framing does a different job than explaining the cut. “CEOs may say these cuts are because of AI, when they really mean to say that they don’t have enough cash flow generation to fund the investments they would like, so they need to trim headcount to free up capital.” Wharton’s Peter Cappelli goes further: “The companies that are laying off are not struggling,” he says. “The answer begins with pressure from investors who always want them to cut headcount.”
Forrester’s J.P. Gownder offers a useful test for telling the difference: “If you are laying people off without a mature, ready-to-go AI agent to do the work, you are not laying off people because of AI. You are laying people off for financial reasons and then imagining that at some future date, AI may be able to do the work.” By that standard, a lot of announced “AI layoffs” are a financial decision wearing a technology label, made ahead of the technology actually being ready to do the job.
If your organization is framing a workforce decision as AI-driven, get specific about what task AI actually eliminated, and be honest if the real driver was cost-cutting or restructuring instead. Data Society’s AI Advisory work helps leadership teams pressure-test that distinction before it becomes a headline, and AI upskilling programs build the fluency to make the call correctly in the first place.
