President Donald Trump is reportedly considering a government review of artificial intelligence models before they are released, reversing his administration’s hands-off approach. Florida Governor Ron DeSantis, from a famously laissez-faire state, has announced his support for legislation moving through Florida’s legislature to protect consumers from rising electricity prices resulting from new data centers. They are responding to increasing concerns about both AI safety and data centers. Yet a new Milken Institute-Harris Poll finds Americans feel left behind on the issue that matters most: jobs. A whopping 80% say they want the government to start preparing workforce transition programs now. Sixty-eight% feel they are navigating the AI transition entirely alone.
Remarkably, business agrees. Eighty-eight percent of business leaders say individual companies cannot solve for AI workforce readiness alone and that a coordinated national response is required. The people deploying the technology and the people living with it have reached the same conclusion. Washington has not caught up.
Those concerns are grounded in data. AI disruption may move faster than anything we have seen before. Researchers have calculated that the speed of adoption of general-purpose technologies has increased with each successive wave — and AI’s low deployment barriers, spreading through software rather than factories, make it faster still. The IMF estimates that over 60% of jobs in advanced economies are already exposed to AI — and unlike prior waves of automation that hit factories and routine tasks, this one lands hardest on cognitive, white-collar work: legal analysis, financial services, software development, administrative functions. A Stanford study found that workers aged 22 to 25 in the most AI-exposed occupations have already seen a 16% drop in employment.
How severe the overall disruption will be remains genuinely uncertain. The Budget Lab at Yale, tracking the labor market monthly, finds its exposure metrics “remain flat, lie within historical ranges, or continue along the trends they were already exhibiting.” The unemployment rate stood at 4.3% in March 2026, roughly where it was before the pandemic. Many economists point to past technological transitions — from the steam engine to the spreadsheet — where employment ultimately rose as new categories of demand emerged that nobody had anticipated. This time may be different. It may not.
But the workers who are already feeling the shift are not in a position to wait and see. Forty-one percent of workers report receiving zero employer AI support in the past year. And while 68% of business leaders expect to be better off in five years, only 27% of workers do. That 41-point gap is a measure of who is absorbing the risk while the country waits for a plan.
The poll also makes clear what workers want: they want help navigating the transition, not a check to sit it out. They express support for specific proposals — including wage insurance during a transition, free community college access to training, notice about upcoming layoffs, and public-private partnerships to coordinate transition.
The problem is that America’s approach to sharing risks and gains was built for an industrial economy — and is not remotely AI-ready.
Two-thirds of the federal budget consists of social insurance — Medicare, Medicaid, Social Security, unemployment insurance — designed for factory workers spending 40 years with the same employer, needing income support only in shorter retirement and for short-term interruptions during their working life when a factory shut down for retooling. These programs don’t offer assistance to the growing share of gig and contract workers during longer transitions, such as training or founding a start-up, and leave them to finance their own health care, retirement, and training with no backstop. The programs need to be modernized along the lines of the administration’s recent effort to extend retirement tax credits to the self-employed.
Other aspects of the tax code need to catch up. As MIT economists Daron Acemoglu and Simon Johnson have documented, current tax disparities place a heavier burden on labor than on capital — effectively subsidizing the replacement of workers with machines even when the automation delivers little additional productivity gain. Labor income also accounts for a higher proportion of federal revenue. If AI further shifts income from labor to capital, the government loses the tax base it needs to fund the very transition programs the public is demanding.
Finally, tech moves too fast for government training programs alone. Business leaders should step up and provide guidance to workers rather than leaving them in the dark. For electricians — badly needed for data centers — apprenticeship programs should be expanded. For jobs needed in the care economy, like teachers and nurses, degree programs may need to be subsidized.
The workers polled do not want to stop the technology. Sixty-nine percent believe AI can create more opportunity than it eliminates. But their families have been burned by the disruptions of globalization and the financial crisis. That failure reshaped American politics in ways still playing out. This time they want to see leadership and a plan that ensures the gains are shared.
The public is watching this moment with considerably less patience. The technology is American. The plan should be too.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
