Lawmakers around the country have taken few steps to stem job losses caused by artificial intelligence, despite widespread concerns about AI-related displacement in the workforce.
Illinois and Tennessee enacted measures this year to protect musicians and performers from unauthorized digital replicas, and a similar New York bill addressing fashion models is awaiting Gov. Kathy Hochul (D). But broader proposals such as a New York “robot tax” measure (A8179) failed to advance and New Jersey bills (S3046 and A1255) that would promote retraining and rehiring workers still await consideration.
AI legislation more broadly has been abundant, targeting deepfake pornography, election misinformation, and discrimination by automated decision-making systems in hiring. But heading off AI-related job losses appears to be a lower priority and a tougher sell politically.
“There are so many threads. Displacement is one piece of a much broader puzzle we’re trying to figure out,” in terms of AI’s potential impacts on workers, said Josh Kellerman, public policy director at the Retail, Wholesale, and Department Store Union. “Policy-wise, I don’t know of anything that gets at that in one fell swoop.”
Major tech companies such as
Then there’s the economic incentives in the annual market for generative AI technologies, forecasted to explode to $1.3 trillion by 2032, up from $40 billion in 2022, Bloomberg Intelligence estimated last year.
Industry-Specific Bills
Many of the statehouse efforts to address worker displacement have been narrow, industry-specific solutions often backed by labor unions looking to protect their members, rather than sweeping attempts to mitigate job loss broadly.
The Teamsters union has urged state lawmakers around the country to require that businesses keep licensed drivers at the wheel of any autonomous commercial vehicles they deploy. No state has enacted the measure.
The California legislature passed such a bill last year, but it was vetoed by Gov. Gavin Newsom (D), a long-time ally of the state’s technology industry. Some state lawmakers are working to send a similar bill to his desk again this year. California is home to
Advocates for fashion models in New York are hoping for better luck with a bill awaiting Gov. Kathy Hochul’s (D) signature. That measure would protect models against the use of digital replicas of their likenesses without their consent—a response to fears that increasingly powerful generative AI is set to largely replace human models.
California lawmakers also are targeting automation of jobs across a handful of other industries, with bills awaiting approval in the final weeks of the 2024 session.
One bill (SB 1446) would restrict the use of self-checkout machines by grocery and retail pharmacy chains. Another (AB 2602) seeks to protect Hollywood actors from being replaced by AI-generated replicas, and a third (SB 1220) would limit the use of AI bots in call centers that contract with the state.
Even in international jurisdictions known for being more worker-friendly and more active in regulating AI, legislative solutions have been scarce. The EU AI Act, a sweeping regulatory scheme for Europe that began taking effect Aug. 1, didn’t include policies aimed at worker displacement concerns, although the European Commission could revisit the topic, said Julia Apostle, an attorney with Orrick, Herrington & Sutcliffe LLP in Paris.
“We know that the Commission is not finished legislating in relation to AI and the text of the AI Act is constructed so as to allow its evolution as the technology evolves,” she said by email.
How Bad Is It, Really?
The public release of advanced generative AI tools such as Google’s Gemini,
But policymakers have struggled to clearly define the scope of the problem and potential job loss timelines.
Employers are predicting widespread adoption of AI technologies over the next five years, with tens of millions of jobs lost—but also tens of millions created—resulting in a net loss of 2% of existing jobs, according a World Economic Forum report published in 2023.
Even tracking how many jobs are currently being lost to AI and automation isn’t easy.
That’s because companies are reluctant to publicly cite AI as the reason for layoffs that they announce, said Colleen Madden Blumenfeld, spokeswoman at Challenger, Gray & Christmas Inc., a job placement agency that tracks and publishes layoff announcement data.
The firm began tracking layoffs attributed to AI adoption in May 2023 and recorded 5,430 job cuts in that category for the calendar year, she said. But in 2024, it has seen no job loss announcements citing AI and instead has recorded over 15,000 attributed to technological updates—the largest number that the agency has seen for that reason since it began tracking it in 2000.
“We believe companies are switching language due to public backlash,” she said via email. “Most of the companies using this language were in industries outside of Tech and traditional huge job creators—Warehousing and Automotive, for instance.”
As another complication for policymakers, figures on job losses don’t capture the disruption caused by AI taking over a portion of a job and possibly leading to a pay decrease for that position, said Ashley Nunes, a researcher on technology in the workforce at Harvard Law School.
Safety net programs such as unemployment insurance and even nascent proposals around universal basic income typically aren’t tailored to help workers who keep their jobs but lose some work and pay, he said.
“This is something politicians and the policy establishment need to be sensitive to,” Nunes said. “Labor economists love to talk about jobs gained and jobs displaced, but what about jobs that are preserved that look very different?”
A coalition of major tech companies is urging state lawmakers to focus their efforts on retraining workers for newly emerging jobs in the industry.
Lawmakers floated a few proposals aimed at workforce development and improving K-12 education around AI this year in Colorado, Connecticut, Maryland, New York, and Washington, and some have proposed developing model legislation that might help states row in the same direction on these efforts.
The federal government has also noted the importance of upskilling, including in non-binding guidance from the House and US Department of Labor that urged employers to follow eight principles in their use of AI tools.
Safety Net, Labor Push
In another approach to the issue, a California bill pending this year—but unlikely to pass after it missed procedural deadlines—would have created an “unconditional benefit income” pilot program, providing $1,000 per month to people unemployed due to automation or AI.
Like the New York robot tax proposal, which would have imposed a tax on employers for each job eliminated and replaced by AI, such a benefit would face the challenge of figuring out which job losses resulted from artificial intelligence, rather than routine cost-cutting.
Disincentivizing companies from replacing workers with AI, such as through a robot tax or by banning the use of driverless trucks, is the wrong policy response, Nunes said.
“If that was the approach we had taken throughout our history, we would still be hand-weaving baskets,” he said. “Having a tax in place to disincentivize the displacement of workers does not do justice to the creativity and tenacity of the human spirit.”
Unions have begun to argue for protections against worker displacement in their collective bargaining agreements, such as requirements that companies must notify and negotiate with worker representatives before deploying new automation technologies, Kellerman said.
“To me that gets at the fundamental problem of technology in the workplace and AI. The silver bullet would be stronger labor laws in this country,” he said.
To contact the reporter on this story:
To contact the editors responsible for this story:
Learn more about Bloomberg Law or Log In to keep reading:
See Breaking News in Context
Bloomberg Law provides trusted coverage of current events enhanced with legal analysis.
Already a subscriber?
Log in to keep reading or access research tools and resources.