The human resources landscape is at a turning point. Recent large-scale surveys of HR and business leaders converge on an uncomfortable reality: few feel confident designing how humans and AI should work together. Everyone has opinions about AI in HR. Almost no one has built the operating model for it.

That gap matters more than you may think because of what’s dissolving underneath it: the job itself. Every reorg of the last 50 years rearranged the same basic bundle of tasks that were required to keep the business running. That bundle is coming apart. Some tasks are now done faster and more consistently by a model than by the person who used to own the whole job. Some still need a person but with a different kind of attention — less execution, more judgment about whether the execution is right. A growing number need both, working together, in configurations no job description ever anticipated.

HR is the function most exposed to that shift, and it’s arriving at a genuine fork in the road.

The current state of HR and possible paths forward

One path is already visible in plenty of organizations: as workforce data, AI governance, and human-agent workflows get more technical, more of HR’s traditional territory quietly migrates to IT and data teams, and HR is left administering what’s left over. The other path is harder and far less common: HR steps forward and claims the mandate for defining how human and AI-driven work actually get combined as the core of what the function does. 

Which path an organization ends up on has less to do with resources than with whether HR closes a set of structural gaps that most functions are still carrying: a gap between operational planning and strategic foresight, between training programs and the pace of actual skill decay, between what employees expect from their organization and how it responds, and between scattered AI pilots and anything resembling scaled impact.

Human resource management

The core mechanics of human resource management (HRM) — hiring, developing, paying, and retaining talent — haven’t changed. What has is the unit those mechanics are supposed to plan against. Workforce planning built around operational capacity (how many people, in which roles, by when) assumed roles held still over the planning horizon. Increasingly, they don’t, which means the discipline that matters now is strategic capability planning: what the organization needs to be able to do, sourced from whichever mix of people, automation, and AI does it best, revisited continuously rather than annually.

That shift runs differently through each core function. For example:

• Recruiting moves from hiring against a fixed job description to hiring against the tasks a model can’t yet absorb, which also means confronting a cost most organizations aren’t pricing yet: as AI absorbs the repetitive, foundational work junior employees have always used to build instincts, the traditional experience-building path thins out, and the leadership bench quietly erodes years before it shows up in a succession plan. 
• Workplace policy has to move at the speed of the tools it’s governing, not the speed of a slow-moving handbook update. Policy written to address new challenges like acceptable use of AI, data exposure, and accountability for an AI-assisted decision that turns out to be wrong must be reviewed continuously rather than written once. 
• Training is shifting from episodic sessions toward continuous co-learning, where people and AI systems adapt to each other inside the daily workflow, with the more durable investment going toward what a model can’t replicate — judgment, ethical reasoning, problem framing — rather than tool proficiency that commoditizes the moment a competitor’s workforce learns the same tool.
• Compensation is pricing two things moving at different speeds: a labor market where AI is repricing entire skill categories and a fairness standard that AI-driven pay-equity tools are making more rigorous to enforce. 
• Retention is now more about trust than compensation. The sharper signal isn’t pay; it’s whether people experienced an AI rollout in their function as something done to them or something they helped shape. 
• Performance management is undergoing perhaps the deepest change of all: once execution runs through an AI agent working alongside a person, the unit worth measuring stops being individual output and becomes the human-agent pairing itself — humans assessed on judgment and orchestration, agents on speed and reliability. 
• Compliance has picked up an entirely new layer, including algorithmic bias, disclosure obligations, and data-use constraints, that moves faster than most legal review cycles were built to catch.

From org structure to flow of work

For years, the default lever for organizational performance was restructuring: delayer, flatten, cut. That lever is running out of room. The bigger opportunity now sits underneath the org chart in how work flows: how many approval layers a decision has to clear, how many handoffs happen before something ships, and how many meetings exist to coordinate work that shouldn’t have needed coordinating in the first place. Middle-management roles have been declining sharply, but flattening the chart without fixing the flow underneath it just produces a flatter version of the same friction.

Organization design

This is where the shift is most structural. Org design is downstream of every assumption above it. If you treat the job as the unit of organization, you’re designing around something that’s already dissolving. If you treat the task as the unit instead — routed to whichever source, human or otherwise, executes it best — the organizing question changes from “how should roles report” to “how should this collection of work flow.”

Most organizations can’t answer that yet because most haven’t dealt with what’s underneath their current structure: years of accumulated handoffs and approval layers, sometimes called organizational or workflow debt. Laying AI over that debt doesn’t remove it. Instead, it executes the same inefficiency faster, with more confidence that it’s been fixed. The stronger path is simplifying the workflow before automating any of it, moving toward what’s often called a skills-based organization, built around capability rather than a fixed chart, and treating structure as something to actively maintain, not revisit once every few years.

The human-and-agent organization

Push the flow argument far enough and a genuinely new organizational shape starts to appear, where a single person leads a team of AI collaborators and output stops being constrained by headcount at all. Growth becomes a function of how well an organization designs human-and-agent teams, not how many people it employs. That’s a vivid image, but the substance underneath it is less about scale and more about governance. Someone still has to decide how those AI collaborators are deployed and remain accountable for it.

Organizations must undergo transformation to fully adopt an effective human-and-agent working model.

HR transformation

HR transformation used to mean the CHRO moving from administrator to strategic partner. The mandate now goes further. To transition to a holistic human-and-agent org model, leaders must guide the enterprise through a workforce transformation touching every function, while dismantling HR’s own siloed operating model at the same time. Neither half can be sequenced after the other, which is what makes this transformation harder than the ones HR has absorbed before.

HR digital transformation

Earlier digital transformation in HR was a systems-of-record problem, moving onboarding and performance tracking off paper. Today, with the rising focus on human-and-agent org models, the shift is more structural. AI agents are active participants in sourcing, screening, and workforce planning now, not passive systems recording decisions made elsewhere. This is a deeper version of digitally enabled HR operations and requires teams to address bigger questions about AI governance, bias, and oversight.

Designing the human advantage

Every lever covered so far — planning, structure, agentic collaboration — eventually runs into the same limiting factor: trust. AI adoption inside a function can scale productivity or it can quietly accumulate what’s sometimes called culture debt — confusion, disengagement, and eroded trust. The organizations pulling ahead are designing the human side of the equation as deliberately as the technical side: expertise that moves to wherever the work is instead of staying locked in a functional silo, work organized around outcomes rather than fixed roles, and learning treated as continuous infrastructure rather than an annual event.

Human resources in business

HR’s core mandate to bring together the right people, with the right capabilities, at the right time hasn’t moved. What’s moved is the planning horizon underneath it: from headcount against a sales forecast to capability against a constantly shifting boundary between human and machine execution.

The traditional case for HR’s importance — hiring quality, retention, compliance, culture — still holds. But the leverage behind it now depends on whether HR has a standing seat in decisions about how work itself gets redesigned. A function consulted after a workforce or AI-adoption decision is finalized is managing the consequences of someone else’s call. A function shaping that decision from the start is doing something categorically more valuable, and that difference is compounding faster than most org charts reflect.

The future of human resources in business

Redesigning how work gets organized touches power structures, career paths, and the operating assumptions every other function relies on. That’s not a scope HR can absorb alone and report back on quarterly. It requires sponsorship only the most senior leadership in a company can provide: actively championing cross-functional, outcome-focused teams, and building enough psychological safety that people are willing to experiment with new ways of working instead of quietly protecting the version of their job they know. This is why AI transformation and workforce redesign now sit at the top of the CHRO agenda — not as HR initiatives to delegate but as enterprise ones to lead from the front.

The organizations that come out ahead over the next several years will be those whose leadership understood early that the job was the wrong unit to be defending and rebuilt around capability, judgment, and flow instead.

That’s the work we do alongside our clients — not handed off and checked on later, but worked through together.

Glossary of human resources terms

Capability planning: A forward-looking approach to workforce planning that focuses on the skills and capabilities an organization needs rather than fixed headcount by role.

Continuous co-learning: An ongoing process in which employees and AI systems adapt to and improve alongside each other within daily workflows, rather than through periodic training.

Employee value proposition (EVP): The overall set of benefits, culture, and experience an organization offers employees in exchange for their skills and commitment.

HR operating model: The structure, processes, and technology through which an HR function delivers its services to the business.

HR transformation: A fundamental redesign of HR’s structure, processes, and strategic role within an organization, often prompted by major shifts like AI adoption.

Human resource management (HRM): The organizational function responsible for hiring, developing, compensating, and retaining employees.

Job architecture: The framework an organization uses to define, group, and level roles based on the scope and complexity of the work involved.

Organization design: The deliberate structuring of roles, reporting lines, and workflows to align how work gets done with business strategy.

Skills-based organization: An organizational model that groups and deploys people based on the skills and capabilities they bring rather than fixed job titles or departments.

Workforce planning: The process of forecasting future demand and organizational needs and preparing the organization’s people, skills, and structure to meet those needs.