Statehouse AI Job-Loss Solutions Range From Punitive to Positive

State lawmakers have tackled the risks of AI-driven job displacement in limited ways, with a division emerging between “carrot” versus “stick” legislative approaches.

Businesses’ adoption of artificial intelligence tools could put up to 100 million Americans out of work within 10 years, Senate Democrats’ labor committee staff said in a report released this month by ranking member Sen. Bernie Sanders (I-Vt.). Economic research is mixed, with some studies predicting AI will create more jobs than it destroys.

Proposed policy responses likewise vary in scope—with some suggesting disincentives such as a robot tax on New York businesses that automate jobs, while others focus on AI skills education and worker retraining. Few have become law, and expansive measures addressing job losses in more than one industry at a time are rare.

The Sanders report called for broadly shoring up worker protections, expanding employee ownership, and making it easier for workers to unionize.

“The agricultural revolution unfolded over thousands of years. The industrial revolution took more than a century. Artificial labor could reshape the economy in less than a decade,” staffers wrote.

But most state and federal lawmakers are still assessing the problem and what policies might best balance worker protections and economic growth, said Niloy Ray, a shareholder with Littler Mendelson PC who advises employers on AI use.

“I don’t think it is fait accompli that AI will suddenly eliminate all these jobs,” Ray said.

There’s been no shortage of state legislative proposals targeting AI-related job losses. Many have sought to stem displacement punitively.

A new Illinois law, for instance, mandates that only licensed therapists can provide psychiatric therapy, limits the ways therapists can use AI for assistance, and carries fines up to $10,000 per violation.

A handful of states this year restricted the use of AI to replace humans, such as community college teachers in Illinois and nurses in Oregon.

Other 2025 proposals aimed to protect California health-care professionals, New York journalists, and Texas schoolteachers from having their jobs automated, according to the University of California at Berkeley Labor Center.

Competing Approaches

The California Federation of Labor Unions pushed for protections of various jobs, such as seeking to limit stores’ reliance on self-checkouts to shield retail cashiers, said Lorena Gonzalez, the federation’s president.

“A lot of our work has been centralized on trying to prevent those job displacements,” Gonzalez said.

But few job loss bills made it through state legislatures, and governors vetoed some of those, including California and Colorado measures to require human drivers in commercial delivery vehicles and another California bill that would have discouraged automation of longshore jobs in the state’s ports.

California Gov. Gavin Newsom (D) also vetoed legislation this year to restrict businesses’ automation of personnel management decisions via the “No Robo Bosses Act” (SB 7)—perhaps the broadest of what Annette Bernhardt calls “human in command” bills.

“This is such a lost opportunity,” said Bernhardt, director of technology and work programs at the UC Berkeley Labor Center. “California had the chance to lead the country in protecting workers in the age of algorithmic management.”

Other states’ AI-related bills aim to use carrots, rather than sticks, to deal with displacement.

Proposals pending in New Jersey would reward businesses that hire workers displaced by automation and provide AI skills apprenticeships.

Utah enacted a 2024 budget bill offering grants for businesses that partner with state agencies on AI career education, while Ohio created a grant for community colleges to offer AI training in 2025 legislation.

Trying to ban companies from automating certain work ignores the history of innovation creating new jobs, rather than mass unemployment, Ray said. The broad adoption of automobiles made New York City’s horse-drawn carriage operators obsolete, he said, but the job of taxicab driver was born.

Offering incentives is likely to be more effective, he said, pointing to a Virginia law requiring that K-12 schools provide AI instruction.

“The quicker we can create people who are adept with AI, both in their existing jobs and in new jobs, the sooner we will be able to spur job growth,” Ray said.

States that seek to penalize AI job displacement will have to contend with President Donald Trump. His administration encouraged states in recent guidance to train workers on AI and has threatened funding cuts to those that regulate AI in ways that could stifle innovation.

Measuring Impact

State lawmakers drafting proposals face the challenge of pinpointing and quantifying job loss due to AI.

A few companies have announced layoffs or reduced hiring while lauding the technology’s cost savings, such as Goldman Sachs Group Inc., Microsoft Corp. and Salesforce Inc. But it’s unknown how many jobs AI will displace or how quickly.

New York Gov. Kathy Hochul (D) attempted this year to get a better view into AI-fueled layoffs, amending the state’s WARN Act notices to require that employers disclose when mass layoffs result from introducing AI or automation.

No employers to date have indicated AI as a reason for layoffs on their WARN forms, state Labor Department spokesperson Christine Buttigieg said.

Businesses have been reluctant to attribute layoffs to AI for fear of public backlash. But that reluctance might be dissipating as executives look to justify their technology investments to shareholders, said Andy Challenger, senior vice president at Challenger, Gray & Christmas, a recruiting firm that tracks corporate layoff announcements.

Challenger’s firm has tracked about 17,000 job losses this year through September where AI or automation was cited as a cause.

“There’s just this huge expectation that companies will get more efficient because of artificial intelligence,” he said.

Lawmakers involved with the State AI Policy Forum hosted by Princeton University’s Center for Information Technology Policy are likely to examine their options on displacement in 2026, said Mihir Kshirsagar, a member of its leadership staff. But it’s hard to predict what comprehensive legislative responses might look like.

That’s “in part because the evidence about the actual effects is still in flux,” Kshirsagar said.