SHRM Research Reveals 23.2 Million US Jobs Already Impacted by AI

SHRM Research Reveals 23.2 Million US Jobs Already Impacted by AI

The Society for Human Resource Management (SHRM) has published research revealing that 23.2 million US jobs — approximately 15.1% of total employment — have at least 50% of their tasks automated or substantially impacted by AI systems. This figure represents the current state of AI labor market penetration, providing a concrete baseline for measuring ongoing workforce AI transformation.

Current Impact Assessment

The 23.2 million figure captures jobs where AI has already automated a majority of tasks, not merely jobs that are theoretically susceptible to future automation. These are positions where AI tools are actively performing more than half of the work previously done by humans — data entry, document processing, customer inquiry handling, financial reconciliation, quality inspection, and administrative coordination.

However, SHRM notes that 63.3% of all jobs include nontechnical barriers preventing complete automation. These barriers include regulatory requirements for human oversight, customer expectations for human interaction, physical manipulation requirements, judgment-intensive decision-making, and organizational policies requiring human accountability. This finding challenges the narrative of wholesale job elimination.

HR Response Patterns

89% of HR leaders expect AI to reshape jobs in 2026 through hybrid human-AI team models rather than simple replacement. Organizations are redesigning roles to allocate routine tasks to AI while preserving human responsibility for judgment, relationship management, and creative problem-solving.

SHRM identifies several organizational response patterns: task redistribution (reallocating automated tasks to AI while expanding human responsibilities in judgment-intensive areas), role consolidation (combining previously separate roles now that AI handles portions of each), new role creation (AI governance, human-AI interface design, AI training), and hybrid role design (creating positions that explicitly combine human and AI contributions).

Skills and Training Implications

The 23.2 million impacted workers need access to reskilling programs that prepare them for evolving role requirements. Only a third of employees have received AI training in the past year, and 72% of employers report difficulty filling AI-related positions. The skills gap is the binding constraint on organizational adaptation.

Workers in impacted roles who develop AI proficiency earn wage premiums up to 56%, creating strong individual incentive for skill development. Organizations that invest in upskilling see measurably higher AI ROI.

The Nontechnical Barriers Protecting Jobs

SHRM’s finding that 63.3% of all jobs include nontechnical barriers preventing complete automation deserves deeper analysis, as it fundamentally shapes the future of work outlook. These barriers fall into several categories that illuminate why the AI labor transition will be transformative rather than eliminative.

Regulatory barriers exist in healthcare, legal services, financial advising, education, and government functions where regulations require human accountability for decisions affecting individuals. The EU AI Act, various US state regulations, and industry-specific rules mandate human oversight for high-stakes decisions, effectively legislating the augmented intelligence model in regulated sectors. These regulatory protections cover approximately 35% of the US workforce.

Relationship barriers protect roles where human trust, empathy, and interpersonal connection are integral to value delivery. Sales professionals, therapists, teachers, clergy, social workers, and healthcare providers derive their effectiveness from human relationships that AI cannot replicate. SHRM’s data shows that organizations attempting to automate relationship-intensive roles experience customer satisfaction declines of 20-40%, driving reversals in automation strategy.

Physical barriers protect roles requiring dexterous manipulation, navigation of unstructured environments, or physical presence — construction, plumbing, electrical work, surgical procedures, emergency response, and personal care. While robotics continues to advance, the cost and complexity of general-purpose physical AI remains far above the economic threshold for widespread deployment in these domains.

Judgment barriers protect roles requiring contextual reasoning in novel situations, ethical deliberation, or integration of tacit knowledge that cannot be codified in training data. Senior executives, judges, diplomats, crisis managers, and creative directors all rely on judgment capabilities that current AI systems approximate but cannot reliably replicate.

Occupational Impact Mapping

SHRM’s research provides granular mapping of AI impact across occupational categories, revealing a more nuanced picture than the aggregate 23.2 million figure suggests. Administrative and office support occupations face the highest task-level automation — approximately 65% of tasks in these roles are automatable with current technology. Customer service representative roles face 55% task automation, primarily for routine inquiries while complex problem resolution remains human-led.

Financial analysis and bookkeeping roles face 50% task automation, with AI handling data aggregation, reconciliation, and pattern detection while humans focus on interpretation, strategy, and client communication. Legal support roles face 45% task automation, with AI managing document review, case research, and contract analysis while attorneys provide strategic counsel and courtroom advocacy.

In contrast, skilled trades face less than 15% task automation, healthcare clinicians face 20% task automation (primarily in diagnostic support and documentation), and creative professionals face 25% task automation (primarily in production and distribution tasks rather than creative conception).

The Hybrid Role Evolution

SHRM documents the emergence of hybrid roles — positions explicitly designed to combine human and AI contributions — as the dominant organizational response to AI’s labor market impact. Rather than eliminating the 23.2 million impacted positions, most organizations are redesigning them into hybrid configurations where AI handles routine components while humans handle judgment, relationship, and creative components.

This hybrid approach aligns with the human-AI team frameworks developed by BCG, Gartner, and Stanford HAI. The most successful hybrid roles share common design principles: clear task allocation between human and AI based on comparative advantage, human-AI interfaces that make AI contributions transparent and evaluable, training programs that develop both AI collaboration skills and the enhanced human skills that hybrid roles demand, and performance measurement systems that capture the combined value of human-AI output.

Goldman Sachs’ projection of a half-percentage-point unemployment increase during the AI transition assumes that the hybrid role evolution proceeds successfully. If organizations fail to redesign roles effectively — either automating too aggressively or failing to integrate AI at all — the unemployment impact could be larger and more sustained than Goldman Sachs projects.

SHRM’s Findings in the Context of Global AI Market Expansion

SHRM’s 23.2 million figure represents the current US impact within a global AI market that reached $196 billion in 2023 and is projected to reach $1.81 trillion by 2030 according to Grand View Research. As AI investment accelerates at compound growth rates, the number of impacted jobs will expand proportionally — SHRM’s figure is a snapshot of an accelerating trend rather than a stable endpoint. McKinsey’s estimate that 40 percent of all working hours will be impacted by AI suggests that the 23.2 million figure could more than double as AI deployment extends beyond early-adopter functions into broader organizational operations.

The World Economic Forum projects 97 million new AI-related jobs by 2025 and 85 million displaced, and SHRM’s US-specific data validates this global projection at the national level. BCG’s finding that AI-augmented workers are 40 percent more productive provides the economic logic driving the job transformation that SHRM documents — organizations redesign roles around human-AI collaboration because the productivity gains justify the organizational change management costs. Stanford HAI reports that AI adoption doubled between 2017 and 2023, and the acceleration of this trend directly correlates with the growing number of impacted jobs that SHRM tracks. PwC’s estimate that AI could contribute $15.7 trillion to global GDP by 2030 depends on the successful management of exactly the kind of workforce transformation SHRM documents — if 23.2 million workers are effectively transitioned into hybrid human-AI roles, they contribute to this GDP growth; if they are displaced without adequate reskilling, the economic potential is lost and the human cost is significant.

SHRM’s role as the world’s largest HR professional society gives its research particular authority with the enterprise leaders who make workforce transformation decisions. When SHRM documents 23.2 million impacted jobs, chief human resources officers treat this figure as an operational planning input rather than an academic estimate. SHRM’s policy recommendations carry weight with employers because the organization represents the professional community responsible for implementing workforce changes, providing a practitioner perspective that complements the analytical perspectives of consulting firms and academic institutions. The organization’s network of 340,000 HR professionals creates both a research channel for understanding employer practices and a distribution channel for evidence-based workforce transformation frameworks that translate research findings into operational changes at the organizational level.

SHRM’s Policy Recommendations

SHRM recommends that HR leaders take proactive steps to manage the 23.2 million impacted positions: conducting AI impact assessments for every role in the organization, developing transition plans for workers in highly-impacted positions, investing in AI skills training programs that prepare workers for hybrid roles, establishing AI governance frameworks that ensure fair and transparent AI deployment in HR processes, and creating feedback mechanisms where workers can report AI tool quality issues and suggest workflow improvements.

The $37.12 billion human-AI collaboration market growth depends on successfully managing this transition. Organizations that handle the workforce impact poorly — through abrupt layoffs, inadequate reskilling, or tone-deaf AI deployment — face employee backlash, regulatory scrutiny, and reputational damage that can offset the productivity gains AI provides.

See our job displacement analysis for comprehensive data, augmented intelligence for market context, future of work for projections, entity profiles, dashboards for tracking, comparisons, and guides for frameworks. For broader labor market context, see the WEF projection of 78 million net new jobs and IDC’s G2000 agent prediction.

SHRM’s HR-Centric Perspective on AI Workforce Impact

SHRM’s research methodology provides a uniquely granular view of AI’s workforce impact because it collects data directly from HR professionals who manage the hiring, training, and restructuring decisions that determine how AI deployment translates into employment outcomes. While economic research organizations model aggregate employment dynamics from macroeconomic data, SHRM surveys the professionals who make individual hiring, termination, and role redesign decisions — providing ground-truth validation of whether macro-level projections are materializing in enterprise-level practice.

The 23.2 million figure represents workers in roles where AI has already materially changed job content, skill requirements, or performance expectations — not workers who have lost their jobs to AI, but workers whose professional reality has been substantially altered by AI deployment in their workplace. SHRM’s data distinguishes between three categories of AI impact: augmentation (workers using AI tools that enhance their existing capabilities, representing approximately 60 percent of the 23.2 million), restructuring (workers whose roles have been redesigned to accommodate AI capabilities, representing approximately 30 percent), and displacement (workers whose roles have been eliminated or significantly reduced due to AI automation, representing approximately 10 percent).

This three-category framework provides a more nuanced picture than binary employed-versus-displaced analyses offer. The augmented majority are experiencing AI as a productivity enhancer that increases their value and often their compensation. The restructured minority are undergoing role transitions that require reskilling but maintain their employment. The displaced minority — approximately 2.3 million workers based on SHRM’s data — face genuine employment disruption that requires institutional support through reskilling programs, unemployment benefits, and career transition assistance.

SHRM’s research also documents the geographic concentration of AI impact within the United States. Technology hub metropolitan areas — San Francisco, Seattle, New York, Boston, and Austin — account for a disproportionate share of both AI augmentation and AI displacement, reflecting the concentration of technology industry employment and early-adopting enterprises in these regions. However, SHRM projects that AI impact will progressively distribute to non-hub regions as enterprise AI deployment expands beyond early-adopting technology companies to mainstream organizations across all industries and geographies, making the geographic pattern observed in 2025 data a leading indicator rather than a permanent distribution of AI workforce effects.

SHRM’s quarterly pulse surveys track how employer attitudes toward AI workforce impact are evolving in real time. The most recent survey found that 72 percent of HR leaders expect AI to create net new roles within their organizations over the next 24 months, compared to 58 percent who held this view one year earlier — reflecting growing organizational confidence that AI augmentation creates more value than it displaces. However, 43 percent of HR leaders report that their organizations lack structured reskilling programs for workers whose roles are being transformed by AI, highlighting the implementation gap between positive attitudes toward AI adoption and organizational readiness to manage the workforce transitions AI creates.

Updated March 2026. Contact info@smarthumain.com for corrections.