Companies are racing to automate work without realizing they may also be automating away the systems that create expertise. That is the workforce challenge emerging beneath today’s AI conversation.
Most debates focus on whether AI will replace jobs. But jobs are the wrong lens for this discussion.
Work has always been made up of tasks, decisions, workflows, and capabilities bundled together under a title and scope. What AI is doing quite efficiently is unbundling that structure and redistributing the work itself. That changes how organizations create value. And it fundamentally changes how leaders should think about workforce strategy.
AI reshapes career ladders
ADP Research has been studying how AI is reshaping work in collaboration with researchers at the Stanford Digital Economy Lab. One of the clearest signals emerging is that the impact of AI is not evenly distributed across organizations. While experienced workers may have anticipated disruption, in practice, many organizations are seeing something more nuanced.
AI performs exceptionally well against structured, repeatable, and rules-based work. Historically, that has often been the work assigned to junior employees. It is where people learn to build repetitions, pattern recognition, and operational judgment.
More experienced professionals tend to operate in environments where context, interpretation, and synthesis matter more. AI can accelerate that work, but it does not fully replace it. In many cases, expertise becomes more valuable in an AI-enabled environment because experienced workers know how to direct, evaluate, and refine machine output.
Experience becomes a multiplier, and that creates a more complicated workforce challenge than many organizations realize.
Erosion at the bottom
The real risk may not be displacement at the top of the organization, but erosion at the bottom.
Business leaders are already beginning to recognize the scale of this challenge as companies invest in upskilling workers to meet the changing nature of job roles and responsibilities. Yet upskilling alone doesn’t answer a more fundamental question: if AI increasingly performs the work where those capabilities were historically developed, how do organizations create the next generation of expertise?
Entry-level work has always served two functions simultaneously: generating operational output and developing future expertise. But as AI assumes more of the repetitive work that once served as a training ground, organizations need new ways to develop the judgment, pattern recognition, and decision-making capabilities that experience has historically created.
When organizations automate too much of that foundational layer too quickly, they risk weakening the long-term capability pipeline enterprises depend on. The challenge is no longer simply workforce automation; it is workforce architecture.
These questions will define the next generation of workforce strategy:
- How do organizations develop judgment when traditional learning pathways disappear?
- How do companies build expertise when repetitive work—historically the training ground for future leaders—is increasingly handled by machines?
- How do enterprises redesign work without unintentionally constraining future talent development?
Answering these questions—not simply deploying AI faster—will separate organizations that build durable talent pipelines from those that quietly hollow them out.
From roles to capabilities
The implications extend well beyond hiring. As work becomes increasingly task-oriented and capability-based, traditional job descriptions become less meaningful than the capabilities individuals can apply across different contexts. Organizations are beginning to shift from static role structures toward capability systems built around a clear-eyed view of what skills exist inside the enterprise, which capabilities are increasing in value, and which activities should remain human-led. The work also requires understanding where AI enhances productivity versus where it introduces risk, and how human and machine work should be orchestrated together.
That transition affects nearly every workforce system: talent development, workforce planning, mobility, compensation, performance evaluation, organizational design, and leadership development. The future of workforce strategy is becoming less about managing roles and more about designing capability ecosystems.
Data as strategic visibility
Historically, payroll and workforce systems were viewed primarily through an administrative lens. But workforce data increasingly provides visibility into how work is performed across organizations: where expertise is concentrated, how workflows operate, where friction exists, and how value moves through the enterprise.
For organizations navigating AI transformation, that visibility becomes strategic.
Companies that understand how work flows through their organizations will redesign faster and more intelligently than those relying on static org charts and outdated role structures.
At the same time, enterprise leaders must recognize that AI adoption inside workforce systems operates differently than consumer AI experimentation.
HR, payroll, benefits, workforce compliance, and employee systems are deeply interconnected, compliance-driven environments. Accountability cannot disappear simply because software becomes more capable.
This is why the future of enterprise AI will not simply be defined by automation. It will be defined by orchestration: where human oversight remains necessary, how accountability is maintained, how workflows are redesigned responsibly, and how organizations preserve trust while increasing efficiency.
While consumer AI optimizes convenience, enterprise AI must optimize accountability. That distinction will shape the next generation of workforce systems.
Redesign work intentionally
Companies succeeding with AI are not treating it purely as a productivity tool. They are redesigning work intentionally. They are using AI to remove friction while elevating human contribution toward high-value activities: interpretation, relationship management, creativity, judgment, and strategic decision-making.
The future of work is not humans versus AI. It is about understanding which systems, decisions, and capabilities are best handled by machines and which remain fundamentally human. The organizations that approach this transition thoughtfully will create stronger, more adaptive workforce models.
The organizations that pursue automation without redesign risk weakening the very capabilities they will depend on most in the future. As leaders rethink work for the AI era, the most important question may no longer be, “What can AI automate?” but “How do we continue developing the human capabilities that organizations will depend on most?” The companies that answer that question well won’t just adopt AI more successfully—they have the potential to build stronger, more resilient organizations for the long term.
Usman “Oz” Khan is senior vice president of ADP Ventures.