6.1 Million US Workers Face AI Displacement with Limited Ability to Adapt

A groundbreaking January 2026 study from the Brookings Institution has identified a vulnerable segment of the U.S. workforce facing a double threat: high exposure to AI-driven job displacement combined with low capacity to adapt to new employment. The research found that 6.1 million workers—approximately 86% of them women—are at significant risk of lasting economic harm as generative AI transforms the labor market.

The study, titled "Measuring US workers' capacity to adapt to AI-driven job displacement," introduces a novel framework that moves beyond simple measures of which jobs AI might affect. Instead, it assesses workers' actual ability to navigate job transitions based on financial resources, age, skill transferability, and local economic conditions—a metric the researchers call "adaptive capacity."

A New Framework: From Exposure to Vulnerability

Traditional analyses of AI's labor market impact have focused on "AI exposure"—estimating which occupational tasks could be performed by AI systems. While valuable, this approach fails to account for vast differences in workers' abilities to withstand and recover from job loss.

The Brookings study, authored by Sam Manning, Tomás Aguirre, Mark Muro, and Shriya Methkupally, addresses this gap by introducing adaptive capacity, a measure built on four key components:

• Net Liquid Wealth: Workers with greater financial savings are better positioned to manage unemployment while searching for new jobs, retraining, or relocating.
• Age: Research consistently shows that older workers (ages 55-64) face greater difficulty finding re-employment and often experience more significant and lasting earnings losses.
• Skill Transferability: The ability to apply existing skills to new roles or industries is critical for occupational mobility. Workers with highly transferable skills face broader re-employment opportunities.
• Geographic Density: Labor markets in more densely populated areas typically offer a greater number and variety of job opportunities, reducing the costs and duration of job searches.

By combining AI exposure data with this adaptive capacity index, the study provides a two-dimensional map of risk, identifying workers who are not only in the path of disruption but are also least equipped to handle it.

The Vulnerable 6.1 Million: Who They Are

The study's most critical finding is the identification of 6.1 million workers—representing 4.2% of the sampled workforce—who face the double burden of high AI exposure and low adaptive capacity. These individuals are at greatest risk of experiencing significant and lasting economic harm from AI-driven displacement.

Demographics and Occupations

This vulnerable cohort is not evenly distributed across the labor market. The analysis reveals stark demographic and occupational concentration:

Gender Disparity: An estimated 86% of these 6.1 million vulnerable workers are women. This reflects the heavy concentration of women in administrative and clerical roles that are highly susceptible to AI automation.

Occupational Concentration: The at-risk jobs are predominantly clerical and administrative support positions—office clerks, secretaries, administrative assistants, receptionists, and medical secretaries. These roles are characterized by tasks that are increasingly automatable by generative AI, coupled with modest wages, limited savings, and skills that may not be easily transferable to other growing sectors.

Geographic Patterns

Contrary to popular narratives that situate technological disruption in major coastal tech hubs, the study finds that the highest concentrations of vulnerable workers are located elsewhere. While cities like San Jose and Seattle have many highly exposed workers, those workers also tend to have very high adaptive capacity.

Instead, the population with high exposure and low adaptive capacity is more prevalent in smaller metropolitan areas, college towns, and state capitals, particularly in the Mountain West and Midwest. Cities identified as having elevated shares of these vulnerable workers include Laramie, Wyoming; Huntsville, Texas; Stillwater, Oklahoma; Springfield, Illinois; Carson City, Nevada; and Frankfort, Kentucky.

This geographic pattern suggests that the negative impacts of AI displacement may be most acute in communities with less dynamic and diversified local economies.

The Resilient Majority—And Why That Matters

The study also finds that of the 37.1 million workers in occupations with high AI exposure, a substantial majority—approximately 26.5 million, or 70%—are in jobs that also have high average adaptive capacity. This group includes professionals like software developers, financial managers, and lawyers.

These workers tend to benefit from higher pay, significant financial buffers, diverse and in-demand skills, and strong professional networks in robust urban labor markets. While their jobs may be transformed by AI, they are generally well-positioned to adapt, retrain, and find new, comparable employment. This finding supports the argument that for many, AI will function more as a productivity-enhancing tool (augmentation) rather than a direct replacement (automation).

However, this resilience among the majority makes the vulnerability of the 6.1 million even more striking—and more urgent for policy intervention.

Implications for Inequality

The findings suggest that without proactive intervention, generative AI is poised to exacerbate existing economic and social inequalities.

Widening Gender Gaps

The disproportionate impact on women in clerical roles threatens to reverse progress made in closing gender-based economic gaps. As AI automates tasks in these female-dominated occupations, it could lead to significant job displacement and downward wage pressure for a demographic group that already faces systemic disadvantages in the labor market.

Broader Brookings analysis indicates that 36% of all female workers are in occupations with high exposure to generative AI, compared to just 25% of male workers.

The Challenge of Worker Voice

The ability of workers to shape the implementation of AI in their workplaces is a critical factor in determining whether the technology leads to shared prosperity or greater inequality. However, a significant "mismatch" exists between the sectors most exposed to AI and those with strong union representation.

Industries with some of the highest AI exposure, such as business, finance, and technology, have exceptionally low rates of unionization (e.g., 1% in finance). This limits the ability of workers in these fields to engage in collective bargaining to secure safeguards, as the Hollywood writers' union successfully did in 2023.

Policy Recommendations: A Three-Pronged Approach

The Brookings research argues that the trajectory of AI's impact is not predetermined. The institution outlines three priority areas for a proactive response designed to mitigate harm and ensure that the gains from AI are broadly shared:

1. Strengthen Employer Practices

Currently, there are few established standards or codes of conduct guiding companies on the ethical and responsible deployment of AI with respect to their workforce. The recommendation is to develop clear guidelines and promote collaborations, such as the partnership between Microsoft and the AFL-CIO, to ensure AI is implemented in a human-centric manner.

2. Enhance Worker Voice and Power

Given the mismatch between AI exposure and union density, strengthening workers' ability to organize and bargain collectively is paramount. This includes federal and state-level labor law reforms that make it easier for workers to form unions and exercise power in their workplaces. Beyond traditional unions, the report suggests exploring innovative models for worker engagement, such as sectoral councils or public-private partnerships.

3. Develop Targeted Public Policy

Policymakers are urged to move beyond broad discussions and enact concrete legislation that addresses the risks identified in the report. The central policy implication is that resources and support should be laser-focused on the workers with the weakest adaptive capacity. This includes:

• Targeted Workforce Development: Creating retraining and upskilling programs specifically for displaced administrative and clerical workers, focusing on building transferable, in-demand skills.
• Modernized Social Safety Nets: Strengthening unemployment insurance and providing financial support, such as wage insurance or emergency savings programs, to help low-wealth workers weather job transitions.
• Place-Based Initiatives: Investing in economic development in the smaller cities and regions identified as having high concentrations of vulnerable workers to create new job opportunities locally.

A Critical Early Warning

The Brookings Institution's January 2026 study provides an essential evolution in the analysis of AI's impact on the future of work. By introducing the concept of adaptive capacity, it shifts the focus from abstract technological potential to the concrete realities facing American workers.

The identification of a 6.1 million-strong cohort of highly exposed, low-capacity workers—overwhelmingly women in administrative roles—serves as a critical early warning. This research makes clear that a passive approach will likely lead to deeper inequality and significant hardship for a specific and vulnerable segment of the population.

The challenge for policymakers and business leaders is to heed this warning and act decisively. By strengthening employer accountability, empowering workers, and targeting public policy to support those least able to adapt, it is possible to steer the AI revolution toward a future of enhanced productivity and more broadly shared prosperity.