Key Takeaways
- AI agents are no longer tools, they are digital workers joining org charts with defined roles, performance metrics, and governance needs
- Neither IT nor HR can manage the human-agent workforce alone; the winning model is shared ownership
- Employee trust is the single greatest determinant of whether human-agent teaming succeeds, and right now, that trust is dangerously low
- HR technology leaders must move from system administration to workforce architecture, owning work design, change management, and agent governance alongside IT
Your Next Hire Doesn’t Have a Pulse.
How HR Technology Must Evolve to Manage a Workforce of Humans and AI Agents
There is a slide making the rounds in boardrooms that would have seemed absurd three years ago. It shows an org chart. Half the boxes have names. The other half have labels like “Recruiting Agent” and “Employee Service Bot.” Both have roles, access permissions, and performance metrics. Only one has people inside.
This is not a concept deck. McKinsey reports that pioneering companies are already expressing their org charts not only in full-time employees but in the number of AI agents deployed across every part of the organization. Workday has built an Agent System of Record to onboard, govern, and manage digital workers using the same platform it uses for human employees. Forrester predicts the top five HCM platforms will offer digital employee management capabilities this year. And Microsoft’s 2025 Work Trend Index, based on 31,000 professionals across 31 countries, identifies the emergence of a new kind of organization: the “Frontier Firm,” structured around on-demand intelligence and powered by hybrid teams of humans and agents.
The human-agent workforce is not a future scenario. It is an operating reality that HR technology leaders must now manage. And the organizations that get it right will not be the ones with the most sophisticated AI. They will be the ones that figure out how to build trust, redesign work, and create governance frameworks that let humans and agents actually collaborate.
This Is Not Another Wave of Automation
It is important to distinguish what is happening now from what came before. Previous waves of workplace technology, robotic process automation, chatbots, copilots, followed a familiar pattern: software assisted humans by completing discrete tasks faster. The human remained in charge. The tool waited to be asked.
AI agents are different. They reason. They plan. They execute multistep tasks across multiple systems without constant human direction. They hand off work to other agents. They escalate to humans when they hit their limits, or at least they should. The Josh Bersin Company describes a three-stage evolution: AI assistants that answer questions, AI agents that execute tasks autonomously, and AI superagents that coordinate multiple agents to manage end-to-end processes. The firm has identified more than 100 agent applications across HR alone, spanning employee services, recruiting, performance management, coaching, learning, and workforce management. Their assessment is blunt: agents and superagents will eliminate up to 30 percent of traditional HR roles, while enabling the professionals who remain to focus on hiring, coaching, and managing the AI infrastructure itself.
This isn’t a marginal shift. Eighty-two percent of business leaders expect to use digital labor to expand their workforce capacity within the next 12 to 18 months, according to Microsoft. CHROs project 327 percent growth in agent adoption by 2027, according to ADP. And Gartner’s latest research finds that while CEO expectations for AI-driven growth remain high, only one in 50 AI investments currently delivers transformational value. The gap between ambition and execution is enormous, and HR technology sits squarely in the middle of it.
The Question Nobody Has Answered: Who Manages the Digital Workforce?
NVIDIA CEO Jensen Huang declared at CES that “every company’s IT department will be the HR department of AI agents.” The line was widely quoted. It was also incomplete.
IT departments already deploy and maintain complex systems, manage identity and access, and handle security. AI agents need provisioning, monitoring, version control, cost tracking, and decommissioning, all traditional IT competencies. On the surface, Huang’s framing makes sense.
But McKinsey Senior Partner Jorge Amar, who advises some of the world’s largest companies on agentic workforce strategy, pushes back. “I don’t think IT will be able to do this alone,” he says. “IT will be critical in enabling the foundational elements, the data stack, the right procurement, the right platform for training and tuning the agents. But the true missing pieces are the business, because nobody can train an agent without knowing the policies and processes intimately, and HR, which will play a key role in pushing the business on what can be done from a hybrid workforce perspective.”
This matters because the hardest problems in human-agent teaming are not technical. They are organizational. Who decides which tasks are handed to agents and which stay with humans? How do you redesign jobs when 39 percent of today’s core skills are projected to change by 2030? How do you maintain team culture when some of your “colleagues” are software? How do you manage the performance of a team that includes both people and algorithms? These are questions that live at the intersection of work design, organizational psychology, and talent strategy, HR’s traditional domain, now operating in radically unfamiliar territory.
The model that Mercer, ADP, PwC, and McKinsey are converging on is shared ownership. IT owns the technical infrastructure: platforms, security, identity, cost management. HR owns the human dimensions: work design, change management, skills development, culture, ethical governance. Both co-own workforce planning, which now must account for human headcount and agent capacity in the same conversation. ADP’s 2026 research puts it directly: “HR leaders’ success will hinge on IT’s expertise in selecting, implementing, and managing complex technologies. At the same time, IT will rely on HR to provide insight into how these tools affect people.”
For HR technology leaders, this means the job has changed. It is no longer sufficient to manage systems. The mandate now includes managing how humans and agents work together, and ensuring the organization is actually ready for it.
The Trust Gap That Could Kill the Whole Thing
The technology is ready. The workforce is not.
Only 46 percent of employees trust AI systems at work, despite over two-thirds using them regularly. Forty-one percent of workers are not sold on the idea of intelligent machines working alongside them. Fears around AI-driven job displacement nearly doubled in 2025, according to KPMG. More than one in four employees have little or no trust in their employer’s ability to deploy AI and automation fairly. And the anxiety is not limited to frontline workers. BCG’s 2025 AI at Work survey found a striking paradox: employees at organizations undergoing comprehensive AI-driven redesign are more worried about job security, 46 percent, than those at less-advanced companies, where the figure is 34 percent. Leaders and managers are far more likely to worry about losing their jobs than frontline employees. The more visible and real the transformation, the greater the anxiety.
People Managing People identifies a new phenomenon they call FOBO: Fear of Becoming Obsolete. This is distinct from job loss anxiety. It is the creeping sense that your skills are degrading in real time, that the window to stay relevant is closing, and that the organization is moving faster than its people can keep up. Their data point is revealing: AI usage jumped 13 percent in 2025, but confidence in using those tools dropped 18 percent over the same period. Workers are complying. They are not trusting.
Shadow AI tells the same story from a different angle. Ninety percent of employees admit to using personal AI tools for work while only 14 percent pay for them. This is not a technology adoption problem. It is a governance and trust crisis. Employees don’t believe their organizations will provide what they need to succeed, so they find their own solutions and hide the evidence.
The implication for HR technology leaders is direct: deploying agents without a trust strategy is deploying agents to fail. Stanford HAI research found that 45 percent of workers favor an equal partnership between workers and AI, and 36 percent want human oversight at critical junctures. PwC’s 2025 Global Workforce survey found that employees with the highest levels of psychological safety are 72 percent more motivated than those who feel the least safe. BCG’s most encouraging finding is that when workers are well-informed and familiar with AI agents, apprehension turns into enthusiasm. The antidote to fear is not reassurance. It is competence, transparency, and genuine involvement in shaping how AI enters the workflow.
This is not IT’s problem to solve. It is HR’s.
Governance as Analysture, Not Afterthought
As agents gain autonomy, governance becomes the load-bearing wall. The consensus across McKinsey, Deloitte, ServiceNow, and Workday is unambiguous: governance is not a barrier to innovation. It is the enabler.
Workday’s Agent System of Record is the most concrete expression of this principle. The ASOR provides tools to onboard agents, define their roles and responsibilities, track their effectiveness, budget and forecast their costs, ensure compliance, and foster continuous improvement. It integrates with Microsoft’s Entra Agent ID to give every agent a unique digital identity, so every action is tracked, attributable, and auditable. Workday CEO Carl Eschenbach describes the vision directly: “As the system of record for more than 10,500 organizations, there is no one better than Workday to manage every part of the workforce, employees, contingent workers, and agents, on our trusted platform.”
The architectural model gaining traction, endorsed by Microsoft and OpenAI, is the agentic pyramid: a base layer of micro-agents with narrow, well-defined functions; a middle layer of tool integrators with precise, limited permissions; and an apex orchestrator that splits tasks, manages failures, and escalates to humans. The design principle is counterintuitive but critical: specialized, constrained agents are far more reliable than monolithic agents attempting to do everything. Organizations that build fleets of focused agents with clear guardrails will outperform those chasing a single super-agent that tries to replace entire departments.
The regulatory landscape reinforces the urgency. The EU AI Act, Colorado’s AI Act effective June 2026, and new California regulations all govern AI in employment decisions. HR technology teams that wait for regulation to force governance will be building it under pressure. Those that build it now will have a structural advantage.
Deloitte adds a practical warning: poorly configured agent interactions can trigger cascading actions and ballooning costs. Organizations need financial operations frameworks specifically designed for agent-driven expenses, tracking token usage, monitoring autonomous resource consumption, and maintaining real-time visibility into the total cost of the digital workforce alongside human labor costs.
The Org Chart Is Changing. The Job Analysture Must Follow.
Mercer makes a compelling argument that traditional job architecture, designed around human-to-human workflows, cannot capture the dynamics of human-agent teaming. Roles must now explicitly include requirements for digital fluency, data literacy, and an aptitude for interpreting AI outputs. Job families and levels should reflect hybrid work models that blend cognitive, emotional, and technical capabilities. New job categories are emerging: Agent Supervisor, Customer Success Orchestration Manager, AI Ethics Specialist.
Microsoft’s data reinforces this. Seventy-eight percent of leaders plan to hire for new AI-specific roles. Top roles under consideration include AI Trainer, AI Data Specialist, AI Security Specialist, and AI Agent Specialist. Twenty-four percent of small and midsize businesses are already considering hiring AI Workforce Managers to lead hybrid human-agent teams.
The most provocative concept in Microsoft’s research is the “agent boss”, an employee who builds, delegates to, and manages AI agents as a core part of their role. Within five years, Microsoft expects this to be a standard responsibility across knowledge work. For HR technology leaders, this means that every learning and development program, every performance management framework, and every career architecture will eventually need to account for an employee’s ability to direct, evaluate, and improve the AI agents on their team.
This is not incremental. It is a redesign of what a “team” means.
Five Things HR Technology Leaders Should Do Now
The path forward is not a wholesale reorganization. It is a deliberate expansion of HR technology’s mandate, from system administration to workforce architecture, executed in stages.
Own the work design conversation. Job deconstruction, breaking roles into tasks and determining which tasks are best performed by humans, agents, or combinations, is fundamentally an HR competency. IT can build the platform. HR must decide what goes on it. If HR technology does not lead this work, it will be done to them, not by them.
Build the agent management muscle. Whether through Workday’s ASOR, ServiceNow’s orchestration tools, or custom platforms, HR technology teams need capabilities for onboarding, monitoring, governing, and retiring AI agents, right alongside human workforce management. Treat this as an extension of what you already do, not a separate workstream.
Close the trust gap before you scale. The data is overwhelming: trust determines adoption, and adoption determines value. Invest in transparency about what agents do and don’t do. Involve employees in designing human-agent workflows. Train people to work with agents, not just around them. And measure trust explicitly, not just deployment metrics.
Formalize the IT partnership. Define clear ownership: IT owns infrastructure, security, identity, and cost management. HR owns work design, change management, skills development, and culture. Both co-own workforce planning. Document this. Resource it. Make it real.
Build governance from the start. Identity management, access controls, performance monitoring, cost tracking, escalation protocols, compliance frameworks, all of this must be in place before agents scale, not after something goes wrong. Governance designed after the fact is damage control. Governance designed up front is competitive advantage.
The Real Question
The human-agent workforce is arriving whether HR technology is ready or not. The boards are committed, 82 percent of CEOs plan workforce reductions enabled by AI within three years. The vendors are committed, every major HCM platform is building agent management into its roadmap. The investors are committed, 2026 budgets are shifting resources from labor to AI across the enterprise.
The only open question is who will shape how this actually works for the people involved. If HR technology leaders treat this as a tools problem, another platform to configure, another system to administer, they will be sidelined by IT departments and consulting firms that move faster. If they treat it as a management problem, a question of work design, trust, governance, and human capability, they will define the most consequential organizational transformation since the adoption of the internet.
The org chart now includes workers without pulses. The question is whether HR technology will manage them, or merely watch them arrive.