By: Dave Zielinski
The job descriptions of front-line and middle managers across industries are about to undergo an extreme makeover. When the next generation of artificial intelligence technology emerges, managers will need to be able to oversee intelligent systems and lead people. They’ll be expected to master change management, navigate complex automation, and assess workflows to determine which tasks are best handled by AI and which still require the human touch.
Agentic AI, in which autonomous agents execute multistep processes on their own, will function as virtual co-workers to human employees in the near future, experts say. The projected ability of agentic AI to set its own goals, conduct high-level reasoning, and learn quickly from mistakes will have vast implications for how organizations recruit and develop managers at all levels.
When agentic AI gains a foothold in organizations, managers in both white-collar and blue-collar jobs will find themselves overseeing not just evolving generative AI (GenAI) tools that produce content in response to prompts, but autonomous agents that can execute complex, multistep processes on their own. In a nutshell, agentic AI will perform tasks rather than just answer questions like current GenAI assistants, which are more passive.
Gartner forecasted that by 2028, one-third of worker interactions with AI will feature the use of autonomous agents and action-based AI models for task completion. The Microsoft 2025 annual Work Trend Index also found that almost one-third of executives said they’re planning to hire managers specifically to oversee hybrid teams of humans and AI agents. The Microsoft study also predicted the imminent rise of the “agent boss,” a worker who builds, delegates to, and manages AI agents to amplify their impact.
Agentic AI vs. AI Agents: Know the Difference
Understanding the new competencies required of managers as agentic AI takes root starts with a base knowledge of the technology. The terms “AI agent” and “agentic AI” are often used interchangeably, said Craig Le Clair, vice president and principal analyst at Forrester, when, in fact, there is a significant difference between them. Le Clair described the distinction this way: AI agents are the applications that perform specific functions, while agentic AI comprises the architecture, models, and systems that underpin the applications.
While the term “AI agent” is often used loosely by technology vendors to describe their AI tools, Le Clair said most so-called agents in operation today lack the true hallmarks of agency: adaptability, learning, and autonomous action, along with the ability to make plans, reason through scenarios, and leverage external tools to achieve goals without human intervention. There is a hierarchy of agents including “worker,” “solver,” and “executive” agents that have varying degrees of autonomy, Le Clair said.
It’s rare to see true agentic AI operating in organizations today, said Emily Rose McRae, a senior director analyst with Gartner. Some vendors claim their tools are agentic when in fact they’re more rudimentary forms of AI, a practice she calls “agent washing.”
“What many people are calling AI agents are really just chatbots,” McRae said. “The kind of agents capable of setting their own goals, learning from their actions, and making decisions with full autonomy aren’t really on the market yet.”
True agentic AI will soon arrive, experts say, and organizations that aren’t preparing their leaders now to manage amid that transformative technology environment will find it difficult to catch up once the wave hits.
“We’re not far from agentic systems,” Le Clair wrote in a recent report he co- authored with Forrester colleagues. “Within the next three to five years, agents will be trusted to control critical enterprise processes. They will also disrupt traditional enterprise apps by replacing their static interfaces and rigid, rule-based logic. Agents will absorb data, learn from it, and dynamically adapt to changing business needs with intelligent, self-evolving solutions.”
Needed: New Managerial Skill Sets
Within HR, agentic AI will complete end-to-end tasks such as creating training videos, cleaning and analyzing job candidate data, or managing an onboarding process. In the broader organization, advanced agents will be used for tasks like handling supplier invoices, obtaining pre-treatment authorization for health insurance coverage, making trading decisions in finance, or managing complete life cycles in software development.
Redefining Managerial Responsibilities
Experts say agentic AI will require more collaborative human-AI relationships in which employees not only oversee AI actions or outputs but treat the AI like a “digital colleague” they can pose questions to or brainstorm with. Organizational charts of the near future will likely include not just human workers but agents stationed throughout a company.
Managers will need new competencies that enable them to oversee intelligent systems, teach human team members how to work alongside AI agents, troubleshoot these tools if there’s problems, and ensure a base level of cybersecurity as AI agents begin to access sensitive company data without human oversight.
“”These tasks will be the responsibility of managers as they shift into becoming managers of automation, not just humans, and will become standard parts of job descriptions,” Le Clair said.
Leading and Coaching People
Sarah Maris, technical learning lead for Udacity, a division of consulting firm Accenture, said that when agentic AI becomes a standard part of the workplace, middle-manager roles will shift away from overseeing workflows and tracking performance to guiding collaboration between humans and AI agents and managing complexity.
Administrative tasks such as assigning work, monitoring progress, and compiling reports will increasingly be handled by AI agents, Maris said, which will allow managers to spend more time coaching employees, supporting growth, and addressing challenges that require human judgment and emotional intelligence.
“Managers’ primary responsibility will no longer be managing the routine but handling the exceptions, which include interpersonal issues, novel problems, and the key decisions that demand human judgment,” Maris said. “Managers will become the essential human checkpoint, shaping how people and AI work together and knowing when to trust automation and when to step in.”
Hiring for What Machines Can’t Do
Like Le Clair, Maris said these new requirements will soon begin to show up in managers’ job descriptions. Requirements like “experience leading a hybrid human-AI team” will become a common qualification.
“”Companies will hire managers based on the very things a machine can’t do: connect with, develop, and truly lead people,” Maris said.
Managers also will need to collaborate with learning and development groups to teach people on hybrid human-AI agent teams new critical thinking skills, experts say.
“The importance of critical thinking will no longer just be for the ‘digital elite’ or white-collar jobs, where much of the focus has been in the past,” Le Clair said. “It also will become key for blue-collar jobs in warehouses and on factory floors where both front-line workers and managers are reviewing outputs or actions of AI-driven tools. For example, a manager might oversee four or five robot janitors. Humans will need the skills to challenge the actions of autonomous agents when necessary.”
Change Management as a Core Competency
Agentic AI also will elevate the importance of another leadership skill: change management. Employees who already have fears about GenAI taking over their jobs may see those concerns escalate as fully autonomous agents are introduced into the workplace. Others may lack confidence in their ability to oversee or work alongside these sophisticated tools. A recent study from Kyndryl, a technology services company, found that 45% of CEOs said most of their employees are still resistant or openly hostile to AI.
“Managers will need to have change management skills embedded as a core competency in their job descriptions,” Le Clair said. “It used to be change management was relegated to HR or geared only toward bigger disruptions like merging with another organization. But agentic AI brings a different type of change that requires a more continual and subtle type of change management. Leaders will need to become skilled in communicating with their teams about how they’ll need to collaborate and interact with AI agents, how AI agents may impact their career paths, and more.”
Le Clair recently delivered a presentation on agentic AI to representatives of 30 government agencies in Washington, D.C. The most frequent question asked concerned change management strategies. “Leaders wanted to know how to talk to their employees about how AI agents would impact their jobs,” Le Clair said.
Rethinking Leadership Development
The disruption that will be caused by agentic AI — and the new skill sets required to manage in more automation-dependent environments — should cause organizations to rethink their leadership development training, experts say.
“The old leadership playbook doesn’t fit the world we’re heading into,” Maris said. “Companies need to take a hard look at how they’re developing managers and make some fundamental changes.”
Rethinking manager development should start with doubling down on teaching what Maris called “critical” human skills.
“Leaders need deeper training in coaching, creating psychological safety, and showing high emotional intelligence,” she said. “These things can no longer just be annual seminar topics, they have to be at the core of leadership development. In an AI-powered workplace, managers’ value lies in their ability to elevate people, not just oversee their tasks.”
Managers across functions also will need technical training in how agentic AI works, Maris said, but the goal shouldn’t be to turn them into data scientists.
“”The objective instead should be strategic AI fluency,” she said. “Leaders need to understand how to collaborate with AI systems. That includes understanding what agentic AI can and can’t do, how to ask the right questions of the technology, and how to recognize when outputs might be biased or flawed.”
Bryan Ackerman, head of AI strategy and transformation for consulting firm Korn Ferry, said agentic AI represents the kind of digital transformation yet unseen by many senior leaders — one that requires a fundamental change in leadership development practices.
“This is happening faster and with more substantive steps in improvement than we’ve seen in previous digital transformations,” Ackerman said. “There are unique elements to this AI transformation that have to become part of an organization’s development plans for its leaders. If executives try to manage this in the same way they’ve managed previous transformations or disruptions, they’ll quickly find they’re moving too slowly.”
Ackerman said executives will be faced with decisions such as deciding what the optimal mix of AI agents and humans in the workplace is and where uniquely human skills will remain a competitive advantage to their companies.
Leadership training tied to agentic AI also should get managers comfortable with being in a state of constant transformation, Maris said.
“The introduction of agentic AI isn’t a moment in time — it marks the beginning of continuous transformation,” Maris said. “Leaders need the skills to guide teams through ambiguity, build cultures of resilience, and normalize constant learning and adaptation.”
Le Clair said another essential component of management development should focus on how to build trust with workers.
“Managers will need to become trust officers in this new environment,” Le Clair said. “The biggest inhibitor to the adoption of AI agents as co-workers will be creating trust in the workforce in these new tools.”