Kevin Frazier
Exactly half of Americans report living paycheck to paycheck, according to a recent Emerson survey. It’s unsurprising, then, that economic anxiety is widespread. Americans are particularly concerned about their savings, their ability to cover medical expenses, and their ability to cover monthly bills. Surely this would result in state legislators prioritizing a job creation agenda—perhaps lower taxes, fewer regulations—and pursuing all methods to increase affordability, right?
The Wrong Policy Approach Penalizes Both Innovation and the Labor Market
As state legislatures return to session, they appear fixated on regulating artificial intelligence (AI), perhaps believing that AI is to blame for the economic woes that are top of mind for voters and perhaps blinding themselves to the technology’s ability to lower health care costs, for example. More than 100 AI bills were signed into law at the state level last year. Early indications suggest that 2026 will see a similar level of AI regulatory fervor. Yet, these bills typically lack a connection to empowering Americans to thrive economically in the Age of AI.
Voters worried about their pocketbooks and family budgets have cause to ask, “Are legislators addressing widespread economic concerns by focusing their legislative attention on AI?”
Many legislators seem to think so. State Senator Scott Wiener—a sponsor of expansive AI regulations in California—recently lamented at the AI+ Summit that “[j]ust telling people, ‘Oh, don’t worry, we’ll retrain you,’ to a 53-year-old accountant whose job has just evaporated,” is a “tough thing.” Perhaps that’s why other legislators, such as Senator Bernie Sanders, have called for a robot tax.
The assumption underlying these and related anti-AI policies is that AI should not be permitted to disrupt existing professions nor to alter business operations. For Sanders and others, the creative destruction of innovation—the process by which America’s economy has evolved and grown—is too chaotic to condone; they’d rather the economy be managed so that changes only occur pursuant to a fixed schedule.
Tragically, this is a recipe for stagnation. If left unchecked, policies intended to freeze today’s jobs in amber will instead freeze the economy.
AI is unquestionably causing economic change, but it is not the source of economic woe. Recall that Americans’ economic concerns predate the introduction of AI. They’ve been battling inflation, high rent, and expensive health care for years—not just since November 2022, when OpenAI launched ChatGPT‑3.5.
Similarly, workers have felt stuck in bad jobs for much longer than the world has been promoting AI tools. The “quits rate”, the percentage of workers voluntarily leaving a job, has been frozen or dropping since November 2021. Finally, for an even longer period, workers have lamented a lack of control over their hours, compensation, and schedule.
Labor is a derived demand. Firms demand more labor based on consumer demand for goods and services. AI, like any other general-purpose technology, is changing what consumers demand and how firms make their goods and services available. In short, AI is transforming the economy, but that’s distinct from destroying the economy.
What a Better Policy Response Might Look Like to Support Innovation and the Labor Market
If legislators want to address the root of the economic concerns shared by so many, they ought to focus less on trying to steer the economy in one direction or another. This is difficult enough when it comes to macroeconomic stabilization, where legislators often overstimulate or crash spending through ill-timed fiscal policy. But it’s even more misguided if attempting to affect the capital-labor split, technology levels, or industrial composition of an ever-evolving economy.
A tax on robots, for example, may decrease economic growth due to lowering productivity. The downstream effects of reduced growth include fewer jobs and lower wages. In general, the idea that policymakers must “do something” about AI is not aligned with a popular demand for a stronger economy.
For policymakers seeking to empower Americans to succeed in the labor market rather than merely pointing a finger at AI, there are two steps forward based on a single, very simple principle: the best way to reduce unemployment is to allow the market to create new jobs.
The reality of a dynamic economy—especially one that moves at the speed of AI progress, which has been and promises to be rapid—is that people need a higher degree of flexibility to find the most valuable and secure job or jobs. A fluid labor market allows several positive and compounding economic benefits. First, workers leave jobs they hate or perhaps are ill-suited for. Second, they then find work that aligns with their skills, schedules, and interests.
These outcomes are positive for firms, too. Disengaged workers sap workplace morale and hinder an office’s productivity. The checked-out employee costs their firms thousands of dollars in lost productivity each year. In a more fluid labor market, these employees will leave sooner rather than later, allowing for better labor matching across the economy. When workers are operating at their highest level of productivity, they do better work for their employer and contribute to the expansion of the US economy. Another benefit is that new firms, such as the startups driving the AI economy, can more easily recruit and retain qualified workers. Armed with better talent, these nascent firms can ensure the country remains at the vanguard of different industries.
To support this, policymakers can take steps that remove barriers and encourage entrepreneurial activity. This benefits not only current AI startups but also longer-term opportunities for business across economic sectors. Startups do not suddenly appear in a garage. Small businesses are not something that falls from the sky. Founders and small business owners launch their bold ideas based on practical education, relevant professional experience, and the promise of economic returns. For those concerned about the jobs AI may consolidate or eliminate, the goal should be to accelerate the discovery of the jobs AI can create. That means helping America’s innovators turn a back-of-the-napkin idea into a growing enterprise that can hire local talent regardless of what industry they are in.
A litany of policy reforms could contribute to this important effort beyond those directly aimed at a particular technology. For example, Congress and state legislatures could consider allowing Americans more flexibility when it comes to accessing savings currently in specialized accounts. These accounts, like FSAs, HSAs, and 529s, could be made fully available for individuals to spend as they see fit for a variety of purposes, including starting a business.
If more general permissive use is not on the table, then a number of exceptions should be granted around small business formation, similar to current exceptions for homeownership. This could lower the need for access to capital and be the difference in whether an individual decides to formally act on their idea for the next big thing. It costs tens of thousands of dollars to launch a small business—why not allow Americans to tap into some of their own funds to get a project off the ground? These and related reforms should all focus on lowering barriers to entrepreneurial activity that could encourage investment not only in AI but also create more entrepreneurial activity in the economy more generally. The jobs of the future will or will not be created based on today’s policies.
Facilitating a freer-flowing labor market is not rocket science (although perhaps rocket science will be more free-flowing in our AI era). A menu of policy options is available to help workers move to opportunities and fully apply their skills. The most obvious reform with respect to freeing up workers is to champion portable benefits at the state and federal levels. The less a worker feels like they need to wrap their arms around their current role, to preserve benefits or to avoid a crazy tax filing season, the more they will be able to apply their skills to the most economically and personally rewarding tasks.
The reality is that these reforms impact more than AI. Every restriction on AI, every barrier to entrepreneurship, and every outdated policy that chains workers to yesterday’s jobs is a vote of no confidence in Americans themselves—in their capacity to adapt, to innovate, and to build something better than what exists today. Innovative policies crafted to welcome and champion dynamism result in Americans having more opportunities and greater autonomy—the sort of liberty at the heart of our republic.
Policymakers must make a choice between paralysis and dynamism beyond just their approach to AI. Broader reforms, rather than protecting specific jobs or even directly supporting a particular new industry like AI, can instead create flexible conditions for new opportunities in possibilities we have not even imagined.
Conclusion
The greatest economic policy is often the one that gets out of the way—not of technology, but of the people who know best what to do with it.
When it comes to AI and the workforce, we should ask a simple question: are we making it easier or harder for someone to start tomorrow what doesn’t exist today? The answer to that question will determine whether America’s economy grows or stagnates, whether workers feel trapped or empowered, and whether the next generation looks back on this moment as the beginning of shared prosperity or the preservation of shared stagnation.