Agentic AI isn’t just joining the workforce. It’s reshaping how organizations plan for it.
AI agents aren’t bound by planning cycles. In an always-on future—where they can forecast, adapt, and align resources in real time—the workforce becomes as dynamic as the business.
Sue Cantrell
Kevin Moss
Russell Klosk
Chloe Domergue
Chris Tomke
Zac Shaw
Agentic artificial intelligence—autonomous systems capable of making decisions and executing tasks with minimal human intervention—is beginning to transform industries and redefine enterprise automation. Deloitte has predicted that, in 2025 alone, 1 in 4 companies currently using generative AI will launch agentic AI pilots or proofs of concept—with adoption expected to reach 50% by 2027.1
Organizations are increasingly considering agentic and other forms of AI as part of their total workforce. But the rise of agentic AI also marks a turning point for workforce planning. Just as these technologies are reengineering not only jobs but also the business processes under which work is performed, we can and should expect them to transform the business processes of workforce planning itself.
No longer a backward-looking exercise done once a year, workforce planning is becoming an always-on process where AI agents can continuously monitor demand and supply, reallocate resources dynamically, and move the function beyond headcount forecasting into more active talent management. This shift prompts organizations to think of workforce planning as the orchestration of a living, adaptive workforce—one where humans and AI agents together anticipate needs, shape capabilities, and deploy resources in real time.
This is the fifth and final shift in our series on the future of workforce planning—and perhaps one of the most transformative shifts (figure 1). Agentic AI does more than just reshape how work gets done. It creates new opportunities for workforce planning, opening the door to more dynamic, continuous, and intelligent ways of aligning talent with strategy.
AI is already rewriting the workforce planning playbook
While agentic AI points to what’s next, AI in its current form is already reshaping workforce planning, helping transform manual processes that, under static planning models, often break down in the face of disruption, leading to misaligned resourcing. AI-powered workforce planning can enable more accurate forecasting of demand and supply based on real-time data, proactive skill gap identification, and the ability to help predict turnover and retention rates. In environments where historical data doesn’t exist, AI can often simulate it by drawing from other sources. And when it comes to short-term or operational planning, it can enable dynamic scheduling in industries like airlines to accommodate daily demand fluctuations and seasonal volatility by integrating multiple data sources to provide a more comprehensive view.
IBM, for example, uses AI in workforce planning to infer employee skills and skill proficiency levels from each worker’s digital footprint, creating a baseline of workforce skills supply. It also employs a machine-learning model to inform salary decisions that are then built into workforce planning models.2 And Walmart uses AI to forecast staffing needs and optimize scheduling, which has led to a 15% reduction in labor costs. By accurately predicting when and where employees are needed, Walmart has been able to reduce unnecessary labor expenses while ensuring that customer service levels remain high.3
AI can also enhance shift and schedule optimization and improve operational workforce planning by analyzing historical data to identify patterns, predict staffing needs, and create more efficient schedules that balance business requirements with worker preferences. NetworkON, for example, estimates that automated scheduling in fleet management operations could save 10% to 15% on labor costs.4 These intelligent systems account for countless variables simultaneously—seasonal demand fluctuations, employee skills and preferences, regulatory requirements, and unexpected changes—delivering schedules that adapt to real-world conditions. Advanced algorithms enable real-time scheduling adjustments in response to unexpected changes in demand or employee availability.5
And for practitioners, these tools are no longer abstract concepts, but a part of daily work. David Boyle, director of workforce planning and analytics at Altria, says he uses AI every day. He leverages AI—including generative AI and AI agents—for a range of tasks: brainstorming potential ideas, conducting executive interviews, and informing and structuring potential scenarios when conducting scenario planning, for example.6
How agentic AI is changing the game
Traditional workforce planning relies on historical data and static models, making it challenging to anticipate rapid changes in demand, workforce availability, or regulatory requirements. This often leads to overstaffing, understaffing, or compliance risks that impact business performance and employee satisfaction.
But agentic AI may open a realm of new possibilities, ushering in a new era of more autonomous workforce planning. Unlike traditional AI models that passively process data and execute pre-programmed instructions, agentic AI possesses the ability to take initiative, make autonomous decisions, and dynamically adapt to its environment. It acts as an independent entity capable of setting goals, strategizing, and executing plans with minimal human intervention.
Agentic AI can analyze vast amounts of workforce data, detect emerging trends affecting talent availability, predict future skill needs, and autonomously generate actionable insights for long-term planning. It can model various workforce scenarios and present humans with various options and recommendations, such as suggesting strategic hiring or training initiatives. Outcomes could also include redesigning jobs by offloading certain tasks to AI agents themselves. Agentic AI can help businesses make proactive, data-driven workforce decisions that improve resilience and agility by continuously learning from real-world data. It can also help democratize workforce planning, moving it from being conducted by a few experts to putting planning, data, and decisions in the hands of every manager (figure 2).
Figure 2
The agentic AI workforce planning shift: From static to dynamic
${optionalText}
Agentic AI does what static models can’t: continuously scan for signals while autonomously forecasting demand, modeling scenarios, and reallocating resources in real time. Unlike traditional systems that stop at reporting, agentic AI keeps learning from fresh data, evolving its recommendations, and sometimes even taking actions autonomously, turning workforce planning into an adaptive, always-on process.
Continuous scanning for signals and patterns
AI agents can continuously monitor and scan for signals and changes in real time, alerting humans or other AI agents when a change occurs.
For example, AI agents can sense attrition in workforces with particular skills, track when contingent workers join or leave the organization and how that impacts skills supply, and identify when workers have extra or declining capacity. They can also pick up signals on worker utilization, productivity, burnout wellness, retention in particular pockets of the organization, and overall work performance and skills. In addition, AI agents can also sense span of control changes, flag siloes or underused skills, spot overlapping or duplicative work, monitor how AI is being used by workers, and flag talent transitions.
AI agents could then feed alerts into planning models and notify leaders when tipping points are reached and workforce plans need to be adjusted accordingly. These signals can also be used to identify trends over time, helping improve predictive models of headcount, skills, and capacity. Workforce plans thus become living systems rather than annual projects, sensing shifts in skills, structures, and tech-human composition in real time.
Autonomous forecasting of demand and supply
Traditional forecasting methods rely on historical data and static models, often struggling to account for real-time fluctuations and external variables like sudden shifts in consumer behavior. But with agentic AI, organizations can more easily detect potential supply and demand mismatches.
Agentic AI can transform demand forecasting by learning from historical patterns, external influences, and real-time data streams to generate precise, location-specific forecasts. Unlike conventional AI models that require manual tuning, agentic AI continuously refines its predictions by recognizing shifts in demand, adapting to new trends, and adjusting forecasts accordingly—live, in real time, and based on real data signals. Beyond demand forecasting, agents can also anticipate future skill requirements and project future workforce needs based on growth, automation, and market dynamics.
They could also reinvent scenario modeling, helping generate or adjust models based on live signals. With agents, organizations can simulate organizational structures, roles, and workforce changes under different scenarios to evaluate readiness and resilience before acting.
When it comes to the supply side of the equation, AI agents can offer powerful new insights that inform decision-making, including:
- Attrition modeling that predicts exits due to retirements, resignations, etc.
- Tracking internal mobility trends by analyzing patterns of promotions, transfers, and career progression
- Identifying internal talent ready to step into key roles to help with succession planning
- Identifying workforce skills with dynamic skills intelligence
- Revealing potential hidden talent through adjacent skills analysis
Dynamic allocation of capabilities and resources
Based on a continuous scan of signals and demand and supply forecasting, AI agents can then suggest interventions and reallocations as needs evolve. Organizations can dynamically match capabilities (in both human and digital workers) to work and adjust workforce strategy (build, buy, borrow, bot) on demand. Worker and AI skills and capabilities are dynamically adjusted in a dynamic capability ecosystem that helps organizations make the shift out of job-based structures.
Consider the simple task of scheduling, for example. Agentic AI can automate scheduling, allowing managers to continuously evaluate constraints, forecast labor needs in real time, and generate optimized schedules that align with business goals and employee preferences simultaneously. Agentic AI eliminates the guesswork and manual effort involved in scheduling, ensuring businesses remain agile and well-staffed by dynamically adjusting to demand fluctuations, scheduling top performers at peak times, reacting to unexpected absences, and managing compliance requirements.
One US hospital network, for example, deployed AI agents to address staff burnout and inefficiencies in workforce management. By deploying agentic AI, the hospital created, automated, and optimized dynamic shift scheduling, resulting in enhanced patient care and reduced clinician workload.7 With autonomous scheduling, fully automated systems handle routine scheduling decisions while escalating exceptions for human review.
Agents will be able to help dynamically orchestrate resources in other ways too, helping inform long-term workforce planning and potential role redesign based on identified patterns. Consider how AI agents are transforming logistics and supply chains, and how similar changes could be applied to the workforce. An event monitoring AI agent, for example, can track when a shipment route is changing, working with an alert AI agent to notify a supply chain planner of the disruption. An impact assessment AI agent could then summarize the implications for the planner and give them suggested options for problem-solving. The planner would then oversee the AI agent in executing the preferred choice, creating a self-healing supply chain.
How might this impact workforce planning? Consider how AI agents could track how this worker chooses to respond to reallocating resources in light of disruptions, as well as how a host of other workers respond as well. Combined data could signal a tipping point in which workforce plans may need to be adjusted, or when a role might be ripe for redesigning. In this case, an organization may decide that it needs fewer supply chain planners due to autonomous AI. For those that they do need, the role may become based on decision-making of exceptions, requiring sophisticated decision management and judgment skills. The organization may even decide that it could combine this updated role across other functional areas beyond logistics and supply chains, with people across functions whose role is to make decisions from options presented by AI agents.
It’s not as far-fetched as it sounds. One global pharmaceutical company is now piloting and experimenting with aspects of this for its supply chain and other functional areas, with the goal of growing without adding additional workers. Consider how AI agents across industries could also identify skills across teams and suggest different team compositions or dynamically match skills to projects and work.
How AI agents move organizations from planning to talent management
With agentic AI, there is little boundary between planning and execution through talent management. After a decision has been made by a human to approve the agent’s suggestion to adjust the workforce mix, for example, another AI agent could initiate a job posting in a recruiting system, or flag potential skills gaps to trigger learning events. It could also interpret job market data and provide suggestions regarding skills to hire for, competitive pay targets, and development priorities accordingly. Likewise, based on continuous sensing of skills and work, agents could automatically and continuously evolve an organization’s skills taxonomy and job architecture.
Technology platforms like Eightfold already offer multiple AI agents as part of their talent management toolbox. Recruiting agents are used to screen candidates, schedule interviews, collect feedback, and even negotiate offers. AI agents tasked with performance management may generate self-assessment reports. And career development agents suggest personalized career growth opportunities based on skills and interests.8
Eightfold, as well as other providers such as Paradox, Seekout, and Maki People, are moving in the direction of building multifunctional agents that do everything from sourcing to candidate support, interviewing, background checking, offer generation, and onboarding.9 Others like Gigged.AI use AI agents to analyze talent, identify what you are missing, and suggest hires that fill the gaps.
Industry spotlight: How agentic AI can address health care worker shortages
Different industries face different challenges. In health care, for example, workforce planners are facing a growing shortage of nurses and other talent, with the World Health Organization estimating a shortfall of 10 million health care workers by 2030.10 Consider how agentic AI could help address these worker shortages.
- Demand alert agents: They sense patterns in demand, such as seasonal flu spikes or public health emergencies.
- Supply alert agents: They sense changes in internal and external talent supply, factoring in attrition signals, mobility patterns, and more.
- Capacity-sensing agents: They sense extra capacity in the workforce, for example, gathering real-time data on clinicians’ workloads.
- Dynamic workforce matching agents: AI agents act as real-time staffing engines, dynamically aligning changing organizational and patient needs with available clinicians based on licensure, specialty, location, capacity, and historical performance, including mobilizing temporary or remote support resources as needed.
- Planning agents: Predictive analytics models, powered by self-learning agents, anticipate workforce needs months or even years in advance by analyzing a combination of historical staffing data, patient population demographics, epidemiological trends, and socioeconomic indicators.
- AI-powered well-being agents: AI well-being monitoring tools track behavioral signals, workload data, and biometric feedback to identify early signs of stress or fatigue. These agents then trigger interventions, such as schedule optimization, alerts to human supervisors, or recommendations for self-care strategies. Over time, such proactive engagement contributes to improved morale, better work-life balance, and lower turnover—key outcomes in a sector facing rising labor attrition.
- Learning agents: AI agents offer clinicians personalized training and upskilling, based on continuously refined models that include improving clinical decision support. AI agents can also help organizations manage clinician certifications, helping keep them up to date.
Making agentic AI work for workforce planning
AI agents are on the horizon, with the potential to transform workforce planning as we know it. We’ve outlined the possibilities of what agentic AI could achieve: dynamic forecasting, continuous learning, and real-time adaptation. Yet the question remains: Can these systems truly deliver on their promise? The verdict is still out, and much depends on how organizations choose to prepare. Success will likely require rethinking processes, building trust and governance, and developing the skills to collaborate effectively with AI agents. With that in mind, here are some key practices organizations can adopt today to position themselves for an agentic future.
- Build internal coordination and consistency. In order to move forward with this level of AI integration, organizations will likely need baseline agreement around multiple issues: governance, operating models, and handoffs, for instance. They should also be clear about which parts of which roles would be handled by agentic AI and which would remain in human hands. Piecemeal discussions are a road to inconsistency (and ineffectiveness).
- Be careful of untrustworthy data. As organizations increasingly rely on AI systems to process, generate, and interpret data, the integrity of these data flows becomes a critical, yet often underestimated, risk factor. Inaccurate data sources can spur decisions based on false assumptions. And while disinformation and misinformation are well-recognized threats, another potential danger comes from “data poisoning”: the contamination or biasing of internal enterprise data pipelines, training sets, or AI outputs, whether by design or neglect. If not addressed, these hidden flaws can silently undermine decision-making quality at scale, degrade model performance, and erode trust, both inside and outside the organization.
- Inform data with contextual understanding. Data gathering and analysis are only part of what goes into effective strategic workforce planning. The other part is understanding the subtext and context around the data, which informs how leaders make decisions. AI could potentially learn this over time, but only through repetitive, meticulous guidance. And even then, there’s always more information than what’s captured in the system.
- Integrate fractured data. To be effective, AI will need to harness data from a variety of sources that are currently fractured, or even contradictory, such as global labor market insights, market job and skills data, current jobs and skills, how workers are spending their time and on what, internal talent and skill mapping, automation assessments, predictive skills intelligence, competitive intelligence, and much more. Organizations should strive to integrate this data and harmonize it to establish a source of truth. Technology vendors covering such a diverse landscape are starting to converge, which should help with this task.
- Consider privacy and ethics. Privacy and ethics are important considerations as well. While it may be tempting to use techniques such as machine learning to predict behaviors, such as the probability that specific individuals will leave the organization, practitioners should consider whether this is a violation of individuals’ rights to privacy.
- Include human oversight, and ensure workers are still practicing critical thinking skills. Significant work still needs to be done before the reasoning ability of agentic AI or the accuracy of its actions reaches the level where it can be trusted to perform mission-critical tasks. Agentic AI can’t think critically and often doesn’t have all the context to make suggestions or take action. And it certainly can’t take accountability for its actions. For this reason, humans need oversight of AI agents. Research suggests that people steered by digital nudges lose some ethical competency and critical thinking, which has big implications for how people use AI agents and assistants.11 Workforce planners should guard against the loss of critical human thinking skills when using AI agents.
- Evaluate the maturity of agentic AI solutions. This is a nascent and emerging technology. Although many vendors are incorporating AI agents into their products, beware of overpromises and thinking that the implementation of AI agents alone will solve the organization’s challenges.
- Align early with information technology, human resources, finance, business and operations, and strategy functions. As discussed in our previous article on cross-functional ownership, adopting agentic AI for workforce planning is a multidisciplinary effort. Engaging cross-functionally before implementation can help ensure that all areas are on the same page from the beginning.
Agentic AI is poised to fundamentally reshape workforce planning—moving it from a static, periodic exercise to an always-on, adaptive discipline. By scanning signals in real time, forecasting demand and supply with greater agility, and managing dynamic resource allocation, AI agents can create new opportunities for competitive advantage. Organizations that prepare now for this shift will likely do more than just keep pace with disruption—they’ll turn workforce planning into a dynamic process that drives sustained business performance.
BY
Sue Cantrell
Kevin Moss
Russell Klosk
Chloe Domergue
Chris Tomke
Zac Shaw
Cover image by: Sofia Sergi