America is racing to adopt AI. But the hardest part may not be building smarter systems—it may be deciding what happens to the people those systems can replace.
The anxiety isn’t science fiction anymore
AI anxiety is getting louder because adoption is becoming measurable. A recent Associated Press story based on a large Gallup Workforce survey describes a sharp rise in how often U.S. workers use AI at work—especially in computer-based, white-collar roles.
At the same time, leaders of major institutions are warning about disruption in plain language. At Davos (WEF 2026), IMF chief Kristalina Georgieva called AI a “tsunami” for jobs and warned young workers and entry-level pathways could be hit hardest, according to The Guardian (and similarly covered by TIME).
What the numbers actually say (and why they still scare people)
Most serious research avoids one clean “X million jobs gone” headline. Instead, it focuses on tasks—the pieces inside jobs. That distinction matters, because task automation can still lead to job redesign, wage pressure, fewer entry-level roles, or consolidation.
A widely cited study by OpenAI and University of Pennsylvania researchers estimated that about 80% of the U.S. workforce could have at least 10% of their tasks affected by large language models, and roughly 19% could see at least 50% of tasks impacted (arXiv paper).
Goldman Sachs economists likewise argued that generative AI could expose the equivalent of 300 million full-time jobs globally to automation and that roughly two-thirds of U.S. occupations are exposed to some degree of AI automation (Goldman report page; Goldman insights summary).
Meanwhile, the World Economic Forum’s Future of Jobs Report 2025 digest and full PDF frame AI as a major driver of workforce restructuring through 2030—alongside large-scale upskilling and role transformation.
So… is it “UBI or everyone’s broke”?
The dramatic framing (“Universal Basic Income or mass unemployment”) sticks because it points to a real fork in the road: if AI boosts productivity, who gets the benefits?
Even without “everyone unemployed,” the U.S. can still end up with something that feels like a crisis for millions: thinner entry-level ladders, more churn, more unstable contracts, and slower wage growth for roles that AI can partially do.
Why “mass insecurity” may be the bigger threat than mass unemployment
A quieter scenario can be just as destabilizing as unemployment: lots of people technically employed, but in lower-security work—while AI productivity gains concentrate in a narrower slice of workers and owners.
- The AI-augmented class: people whose jobs become faster, higher-paid, more influential.
- The AI-replaced class: people pushed into lower-wage work, unstable gigs, or repeated retraining cycles.
That split is part of what fuels renewed interest in UBI and “guaranteed income” models—even among city governments and local coalitions.
What U.S. evidence suggests about giving people cash (and work incentives)
The U.S. doesn’t have a national UBI, but it does have real-world tests of unconditional or near-unconditional cash programs. Two cases are frequently cited in the UBI debate:
1) Stockton’s guaranteed income experiment (SEED)
The Stockton Economic Empowerment Demonstration (SEED) provided $500/month for 24 months to a group of residents. SEED’s published findings reported improvements in financial stability and wellbeing—and reported increases in full-time employment among recipients compared with the control group (program site; findings PDF).
2) Alaska’s Permanent Fund Dividend
Alaska’s Permanent Fund Dividend is not a full UBI, but it’s a long-running, broadly distributed annual cash dividend. A study associated with the National Bureau of Economic Research found no effect on employment overall and a modest increase in part-time work (NBER paper page).
To see how widespread U.S. guaranteed-income experiments have become, the Stanford Basic Income Lab’s Guaranteed Income Pilots Dashboard and Mayors for a Guaranteed Income track dozens of pilots and public initiatives.
If not UBI, what else could the U.S. realistically do?
In practice, the U.S. may move via a mix of policies—some UBI-adjacent, some not. Here are the options that show up repeatedly in mainstream labor-market planning:
1) A bigger, simpler income floor (not necessarily universal)
Expanded tax credits, wage subsidies, or targeted guaranteed income can function like a “partial UBI” without committing to universal payments. This path is politically more plausible in the near term—and aligned with how most U.S. pilots are structured today.
2) Massive retraining paired with real hiring pipelines
Reskilling fails if there aren’t enough jobs at the end. The WEF report emphasizes upskilling as a top employer strategy in response to AI disruption (see the WEF section on workforce strategies: WEF workforce strategies chapter).
3) Job redesign instead of job removal
This is the optimistic model: AI does repetitive tasks while humans handle higher-value work. It’s real in some sectors—but it depends on management decisions, competition, and incentives.
4) A new deal on how productivity gains are shared
Shorter workweeks, profit-sharing, stronger wage growth, new social insurance—these are all ways to distribute AI’s productivity gains more broadly. In other words, AI can create abundance, but abundance doesn’t automatically distribute itself.
The blunt takeaway
You don’t need to predict the exact number of jobs AI will eliminate to act. What’s increasingly hard to ignore is the combination of (1) rapid adoption, (2) high task exposure across the workforce, and (3) credible warnings that entry-level pathways could shrink.
Whether the answer becomes UBI, a targeted income floor, a job guarantee, or something new, the core question is the same: if AI replaces a large share of human labor, what does survival look like when a job is no longer the default?
Sources
- Associated Press — “How Americans are using AI at work, according to a new Gallup poll”
- The Guardian — IMF chief warns of AI “tsunami” hitting jobs, with young workers at risk
- TIME — Davos 2026 interview/coverage with IMF’s Kristalina Georgieva on AI and jobs
- arXiv — “GPTs are GPTs” paper (Eloundou et al., 2023)
- Goldman Sachs Global Investment Research — “The Potentially Large Effects of AI on Economic Growth”
- Goldman Sachs — “Generative AI could raise global GDP by 7%” (summary)
- World Economic Forum — Future of Jobs Report 2025 (digest)
- World Economic Forum — Future of Jobs Report 2025 (PDF)
- SEED (Stockton Economic Empowerment Demonstration) — official site
- SEED — Findings summary PDF (first-year analysis)
- NBER — “The Labor Market Impacts of Universal and Permanent Cash Transfers” (Alaska dividend)
- Stanford Basic Income Lab — Guaranteed Income Pilots Dashboard
- Mayors for a Guaranteed Income — coalition site
