The economic impact of artificial intelligence on American workers has crossed a threshold that transforms it from a subject of academic projection to a lived political reality. The Bureau of Labor Statistics estimates that AI-related automation contributed to approximately 2.8 million job displacements in the United States during 2025 — a figure that represents both a significant acceleration from previous years and a dramatic undercount of the technology’s full labor market impact. The workers losing their jobs to AI are not hypothetical future victims of a coming transformation. They are current constituents whose experiences will shape the politics of the 2028 election and whose circumstances demand policy responses that American institutions have been slow to develop.
The Displacement Data: What We Know and What We’re Missing
Understanding AI’s impact on employment requires acknowledging the severe limitations of available data. The Bureau of Labor Statistics’ displacement figures, while the most authoritative available, capture only a fraction of AI’s labor market impact. The BLS methodology counts workers who lost jobs due to business closures or position eliminations but does not capture several important categories of AI-driven employment change.
Workers whose hours have been reduced because AI systems handle tasks they previously performed full-time do not appear in displacement statistics. A customer service team reduced from twenty agents to eight — with the remaining eight handling escalations that AI cannot resolve — registers as twelve displaced workers, but the eight retained workers who now have diminished responsibilities, reduced overtime, and weakened bargaining positions are invisible in the data.
Workers who remain employed but at lower wages because AI has increased the supply of labor for their remaining tasks are also invisible. A graphic designer who previously earned $80,000 annually but now competes with clients who can generate acceptable designs using AI tools and consequently accepts $55,000 for more limited services has not been displaced — but their economic situation has been materially damaged by AI.
New graduates who cannot find employment in fields that historically absorbed entry-level workers represent another uncounted category. The collapse of entry-level positions in content writing, basic legal research, financial analysis, and customer service has not displaced anyone from an existing job, but it has closed pathways into professional careers that previous generations relied upon.
When these broader categories are considered, estimates of AI’s total labor market impact are substantially larger than the headline displacement figure. Research from the Brookings Institution suggests that approximately 11 million American workers experienced material negative employment effects from AI in 2025, including displacement, hours reduction, wage compression, and blocked career entry.
Sector-by-Sector Analysis
AI’s impact on employment varies enormously by sector, with some industries experiencing rapid transformation while others remain largely unaffected. Understanding these sectoral dynamics is essential for effective policy response.
Customer Service and Call Centers
The customer service sector has experienced the most visible and rapid AI-driven employment reduction. AI chatbot and voice assistant systems now handle an estimated 65% of routine customer interactions across major corporations, up from approximately 25% in 2023. The U.S. call center workforce has declined from approximately 2.9 million in 2023 to approximately 2.1 million in early 2026, with projections suggesting further decline to approximately 1.5 million by 2028.
The remaining human customer service roles have shifted dramatically upmarket. Workers now handle primarily complex, emotionally sensitive, or high-value interactions that AI systems cannot manage effectively. This shift has improved average wages for retained workers but has eliminated the vast majority of entry-level positions that historically provided employment for workers without college degrees, including significant numbers of workers in rural areas where call centers were among the few available sources of stable employment.
Content Creation and Media
The creative and content industries have experienced what might be described as a quality-stratified impact. AI has not eliminated demand for human creative work at the highest quality levels — premium journalism, sophisticated marketing strategy, complex creative direction — but it has devastated the market for competent-but-routine content production.
The freelance content writing market has contracted by an estimated 60% since 2023, measured by the total dollar value of assignments available on major freelance platforms. SEO content, product descriptions, social media posts, basic reporting, and marketing copy — categories that previously supported hundreds of thousands of freelance and staff writers — are now predominantly generated by AI systems with minimal human oversight.
The impact extends beyond writing. Stock photography revenues have declined sharply as AI image generation replaces purchased photography for many commercial applications. Basic graphic design, video editing, and audio production have all experienced significant demand reduction for entry-level and mid-level work.
Legal Services
The legal profession has experienced AI’s impact primarily through the automation of tasks that previously required junior associate and paralegal labor. Document review, contract analysis, legal research, and regulatory compliance checking — tasks that collectively represented a substantial portion of law firm revenue and a critical training ground for young lawyers — are increasingly performed by AI systems at a fraction of the cost and time.
Large law firms have reduced entry-level hiring by an estimated 25-35% compared to pre-AI projections, though the impact has been partially masked by the profession’s cyclical hiring patterns. The more significant impact may be on the structure of legal careers. The traditional pyramid model — many associates at the base generating revenue through high-volume document work, with a narrow partnership track above — is becoming unsustainable as AI eliminates the economic rationale for the base of the pyramid.
Financial Services
AI’s impact on financial services employment has been concentrated in middle-office and back-office functions: data entry, transaction processing, compliance monitoring, basic financial analysis, and report generation. Major financial institutions have announced workforce reductions ranging from 10% to 30% in these functions, with implementation timelines extending through 2028.
Trading floors have experienced less displacement than some early predictions suggested, primarily because the shift from human to algorithmic trading was substantially complete before the current AI wave. However, AI is now enabling a second wave of automation that affects risk analysis, portfolio construction, and client reporting functions that survived the first round of fintech automation.
Healthcare
Healthcare has been a notable exception to the displacement pattern, at least at the direct patient care level. AI has not reduced demand for doctors, nurses, or allied health professionals, and is unlikely to do so before 2028. The shortage of healthcare workers in the United States is severe enough that AI systems augmenting clinical capacity are generally experienced as welcome support rather than threatening competition.
However, healthcare administrative functions have experienced significant AI-driven automation. Medical coding, billing, prior authorization processing, and appointment scheduling are increasingly AI-automated, with employment in healthcare administration projected to decline by 15-20% by 2028. This administrative automation has the potential to reduce healthcare costs, but the employment impact falls disproportionately on workers in administrative support roles that have historically provided stable employment without requiring clinical credentials.
The Policy Response: Too Little, Too Late?
The American policy response to AI-driven labor displacement has been characterized by acknowledgment of the problem, development of pilot programs, and a persistent gap between the scale of the challenge and the resources deployed to address it.
Federal workforce retraining programs have received increased funding but remain modest relative to the scope of displacement. The Workforce Innovation and Opportunity Act (WIOA), the primary federal workforce development program, was funded at approximately $6.3 billion in fiscal year 2026 — an increase from previous years but still less than the cost of a single frontier AI training run. The mismatch between the scale of the technological transformation and the scale of the policy response is stark.
The effectiveness of existing retraining programs is also contested. Evaluation data from WIOA-funded programs suggests that participants experience average earnings increases of 10-15% compared to comparable non-participants, but these gains are concentrated among younger workers with existing technical foundations. Older workers, workers in rural areas with limited local labor markets, and workers without postsecondary education achieve substantially smaller gains and face higher dropout rates from training programs.
Community college systems have emerged as a critical infrastructure for AI-era workforce development. Several states have launched targeted programs that combine technical training in AI-adjacent skills (data analysis, prompt engineering, AI system management) with credentials that employers recognize. These programs have shown promising early results but face capacity constraints — community college enrollment and staffing have not yet scaled to match demand for retraining services.
Universal Basic Income and Income Support
The debate over universal basic income has intensified as AI displacement has accelerated, but political support for UBI remains insufficient for federal enactment. However, more targeted income support proposals have gained traction. Several legislative proposals would create “AI displacement insurance” — a program modeled on trade adjustment assistance that would provide extended income support and retraining benefits specifically for workers displaced by AI automation. These proposals have attracted bipartisan interest, with support from both labor-aligned Democrats and some Republicans who see targeted AI displacement programs as preferable to broader expansion of social safety net programs.
At the state level, the policy response has been more varied. California has enacted the most comprehensive state-level AI workforce protection law, requiring companies to provide 90 days’ notice before AI-driven layoffs affecting more than 50 employees and contributing to a state AI retraining fund proportional to the number of positions automated. Washington, New York, and Illinois have enacted similar but less comprehensive measures.
The 2028 Election and the AI Economy
The political implications of AI-driven workforce displacement will be a major factor in the 2028 presidential election. The workers most affected by AI displacement — call center workers, content creators, administrative professionals, and junior knowledge workers — represent a new political constituency that does not map neatly onto existing partisan alignments.
These workers are disproportionately concentrated in suburban and exurban communities that have been swing demographics in recent elections. They are generally educated, have been solidly middle-class, and are experiencing economic displacement for the first time in their careers. Their political allegiance will be contested by candidates from both parties, and their responses to AI will be shaped by direct personal experience rather than abstract ideological positioning.
The economic data suggests that AI displacement will intensify between now and November 2028. Projections from multiple research groups estimate that cumulative AI-related job displacement in the United States will reach 8-12 million positions by late 2028, with additional millions of workers experiencing hours reduction, wage compression, or career pathway disruption. These figures, if realized, would make AI displacement one of the most significant economic forces shaping voter sentiment in the 2028 cycle.
Candidates who can articulate credible AI workforce policies — combining support for workers displaced by automation with strategies for maintaining American competitiveness in AI development — will have a significant advantage. The political challenge is real: voters simultaneously want protection from AI displacement and access to the economic benefits of AI adoption. Policies that appear to slow AI development in the name of worker protection risk backlash from voters who see AI as essential for American competitiveness. Policies that embrace AI acceleration without adequate worker support risk alienating voters experiencing displacement.
What Effective Policy Would Look Like
Addressing AI-driven workforce displacement at the scale currently emerging requires policy responses that are substantially more ambitious than anything currently enacted or seriously proposed. Several principles should guide policy development.
First, the speed of displacement demands rapid response mechanisms. Traditional workforce development programs operate on timelines of years — but workers who lose their jobs to AI this quarter need income support and retraining access this quarter. Programs must be designed for speed of enrollment and delivery, even at the cost of the comprehensive assessment and planning processes that characterize existing programs.
Second, retraining programs must be realistic about outcomes. Not every displaced worker will become an AI engineer or data scientist. Effective programs must include pathways into roles that are genuinely in demand, that match the aptitudes and constraints of the workers being served, and that offer wages comparable to the positions that were lost. Programs that train workers for oversaturated fields or that promise career transitions that are unrealistic for most participants waste resources and erode trust.
Third, income support during transition must be adequate to prevent financial crisis. Workers who face foreclosure, medical debt, or family dissolution during a six-month retraining program are unlikely to complete that program successfully. Transition support must cover the actual costs of maintaining a household, not just a fraction of previous earnings.
Fourth, the costs of workforce transition should be distributed across the economic actors who benefit from AI adoption. Companies that reduce labor costs through AI automation capture the economic gains of displacement. A portion of those gains should fund the transition costs that displaced workers bear. Tax policies, mandatory retraining contributions, or AI displacement insurance mechanisms can accomplish this distribution without creating barriers to beneficial AI adoption.
The 2028 election will be the first in which AI workforce displacement is a first-tier economic issue. The quality of the policy debate — and the credibility of candidates’ workforce proposals — will substantially influence both the election outcome and the economic trajectory of millions of American workers.
Employment data in this analysis draws on Bureau of Labor Statistics reports, Current Population Survey microdata, and proprietary analysis of job posting databases. Displacement projections incorporate sector-specific automation probability estimates and are updated quarterly. Last update: Q1 2026.