Is AI Adoption Reinventing the Workforce?

As AI takes on more customer service tasks, CX leaders must redefine what people do best, from complex problem-solving and judgment to empathy and relationship-building.

by Execs In The Know

Artificial intelligence (AI) adoption has entered a new phase. For years, the conversation centered on what AI might eventually do. Now, the more consequential question for customer experience (CX) leaders is what happens to the work around it.

As AI takes on more routine interactions, summarizes customer conversations, recommends next-best actions, identifies patterns, and increasingly executes tasks on its own, customer service organizations are rethinking the basic architecture of their workforces. The question is no longer simply how many people are needed to handle a given volume of contacts.

It is becoming: What should people be doing when AI is doing more of the work?

That is a much bigger question.

It touches job design, skills, compensation, workforce planning, career paths, organizational structure, leadership, and the relationship between CX and human resources. It also raises an uncomfortable possibility: If organizations automate the work that traditionally helped people develop experience, judgment, and expertise, how will they build the next generation of experts?

This is where workforce reinvention becomes less about implementing another technology and more about redesigning the business.

Gartner research shows how quickly that shift is happening. In a recent survey, 85% of customer service and support leaders said they are expanding human agent responsibilities as AI reduces contact volume and shifts work toward higher-value tasks.¹ At the same time, 80% said they are under pressure to make workforce changes because of AI.

The workforce is not disappearing. It is changing shape. And for CX leaders, that may be the more important transformation to get right.

From Job Elimination to Job Redesign

The easiest way to think about AI’s impact on the contact center is to imagine a shrinking pyramid: fewer agents at the bottom, more automation at the top. The reality is likely to be more complicated.

AI can absorb large portions of work that once consumed an agent’s day: authentication, information retrieval, simple troubleshooting, order-status questions, appointment scheduling, summarization, after-call work, knowledge searches, and increasingly complex transactional tasks.

That does not necessarily make the human role less important. It changes the nature of the role. An agent who once spent most of the day answering repetitive questions may increasingly spend that time handling exceptions, resolving emotionally charged situations, navigating complex policies, protecting customer relationships, or making decisions where context matters more than speed.

The value of the role moves up the ladder. Gartner found that only 20% of customer service leaders reported reducing agent staffing because of AI, while 55% reported maintaining staffing levels even as they handled higher customer volumes.²

That distinction matters. If the goal is simply to remove labor, AI becomes a cost-reduction exercise. If the goal is to redesign work, AI becomes an opportunity to change what the workforce can accomplish.

“One of the biggest misconceptions about AI is that it will replace the human element in customer service. In my view, AI actually elevates the importance of distinctly human skills. At Michael Kors, we’ve invested in AI through initiatives like Mia, our virtual assistant, to handle routine, transactional interactions. Customers don’t necessarily want to speak with a person to check an order status or find a return label. They want a fast, accurate answer. AI is incredibly effective in those moments,” explains Ebrahim Hyder, Vice President of Customer Service at Michael Kors. “Where humans become even more valuable is when emotion enters the conversation. Empathy, judgment, relationship-building, and the ability to interpret nuance are skills technology still struggles to replicate authentically. A customer calling because a gift didn’t arrive in time for a special occasion isn’t looking for a process. They’re looking for understanding.”

The question leaders should be asking is not: how many agents can AI replace?

It is: what work should our people be doing once AI removes the work that no longer requires a person?

That question leads somewhere much more interesting.

The Agent May Become an Orchestrator

The next generation of CX work may require employees to manage systems as much as they manage individual customer interactions. Consider what happens when AI agents can investigate an issue, retrieve information from multiple systems, recommend a resolution, initiate a transaction, and escalate when they reach the edge of their authority.

The human employee is no longer simply the person answering the customer. They become responsible for ensuring the customer gets the right outcome. That requires a different set of capabilities.

Employees will need to understand how AI systems behave, recognize when an AI recommendation is unreliable, interpret data, investigate exceptions, manage escalations, understand business rules, and make judgment calls when the answer is not obvious. They will also need to know when not to trust the machine.

This distinction matters because AI can create new work even as it eliminates old work. Someone has to monitor performance, validate outputs, correct errors, improve knowledge, manage exceptions, maintain context, and determine where the system’s authority should stop.

IBM’s 2026 research found that 80% of CHROs believe AI adoption creates “invisible” work, including validating recommendations, fixing mistakes, providing context, and managing exceptions.³ That invisible work could become a defining characteristic of the AI-enabled workforce. The best organizations will make it visible, define it, measure it, and build roles around it.

The Skills That Come Next

If the work changes, the skills have to change with it. Data literacy will matter. So will AI literacy. But neither should be confused with simply knowing how to use a chatbot.

CX organizations will need people who can connect customer signals to business decisions. They will need employees who can recognize patterns, frame problems, question assumptions, manage increasingly complex workflows, and understand the consequences of an automated decision.

They will need strong communicators and people with sound judgment. They will need employees who can work across functions because the customer journey rarely respects the boundaries of a contact center, product team, technology organization, or finance department. Some of the emerging roles may have distinctly new titles. Others may look surprisingly familiar.

AI operations specialists. Conversation designers. AI trainers. Automation analysts. Knowledge managers. Workforce strategists. Customer journey orchestrators. AI governance specialists. Service designers.

But the larger change may happen inside existing roles.

A workforce manager may increasingly manage a blended workforce of humans and AI agents. A quality leader may evaluate AI interactions alongside human ones. A training leader may teach employees how to collaborate with AI rather than simply how to follow a process. A contact center leader may become responsible for the performance of an entire service ecosystem rather than a defined population of agents. That is workforce reinvention in practice.

Fewer Agents, Different Work

Another shift leaders need to confront is that the workforce may eventually shrink in some areas, but that doesn’t mean the remaining jobs simply become better versions of today’s jobs. They may be fundamentally different.

Agentic AI changes the equation because AI can move beyond answering questions and begin taking action. Gartner also found that 58% of customers who use GenAI have used it to complete a task on their behalf, rising to 74% among B2B customers.⁴

That points toward a service environment where AI agents do not merely assist human agents. They perform portions of the customer journey themselves. A customer might ask an AI agent to change a reservation, resolve a billing issue, find a replacement product, submit documentation, or coordinate an appointment. The AI may handle the entire interaction. Until it cannot. And that “until” is where the human workforce becomes strategically important.

That handoff is already a weak point. In our recent CX Agent Insights: Perspectives from the Frontline report,⁵ agents ranked escalation procedures as the second-biggest area needing improvement in their operations (18%), behind only technology.

A human may step in when the situation is unusually complex, emotionally sensitive, financially significant, ambiguous, or simply outside the AI’s authority. The employee may have fewer interactions overall, but each interaction may carry more consequence. That has implications for staffing models.

If a contact center has fewer routine interactions but a higher concentration of complex ones, traditional productivity measures may stop telling the whole story. Average handle time could become less meaningful when the objective is thoughtful resolution.

Agents already seem to know this. When asked to rank what should matter most in measuring their performance, frontline agents put customer satisfaction and first-contact resolution first, and average handle time near the bottom.⁶

Agent occupancy could tell a misleading story if employees spend time preventing escalations, solving systemic problems, or supporting AI-driven workflows.

The workforce model itself may need to change. And so may the way employees are paid.

The Compensation Question

This is one of the least discussed aspects of AI-driven workforce transformation. If the job becomes more complex, should compensation remain anchored to the old job description? An employee who once handled straightforward service requests may now be expected to investigate difficult cases, exercise greater judgment, manage AI outputs, understand multiple systems, and protect high-value customer relationships.

That is a different job, even if the title on the organizational chart has not changed. Workforce reinvention therefore raises questions about pay scales, career progression, job levels, performance metrics, and incentives. It also creates an opportunity to make CX careers more attractive.

For years, the contact center has struggled with perceptions that frontline service work is repetitive and transactional. AI could remove some repetition and create pathways to more sophisticated work. But that won’t happen automatically. Leaders have to design it.

The ambition is already there. Two-thirds of frontline agents say they plan to stay and grow in the field, but fewer than half say their organization offers a clear path to do it.⁷

• Only 47% of agents say their organization offers clear career paths and advancement, yet 67% plan to stay and grow in the field (p. 11).
• Career growth is the top engagement driver other than pay (27%). Recognition is near the bottom (6%) (p. 15).
• 45% of respondents have five or more years in the role (p. 7).
• Pay came up repeatedly when agents were asked what would improve their work-life balance (p. 14). One agent said: “Make agents happy (rewards and pay more) to make customers happy.”

They decide what an entry-level employee learns first, what experiences help that employee develop judgment, which skills lead to advancement, and what a career in CX looks like when AI handles much of the routine work.

That is a workforce strategy question, not a technology question.

Who Owns the Reinvention?

Which brings us to HR. How involved is HR in your AI strategy today? That question deserves more attention than it receives. According to IBM’s 2026 CHRO study, 46% of organizations don’t involve the CHRO when defining AI strategy.⁸ Only 26% clearly define work across human-led, AI-assisted, and AI-executed activities.

Those numbers suggest a significant gap between technology strategy and workforce strategy.

Reinventing work is difficult without involving the people function responsible for job architecture, skills, compensation, career development, organizational design, and change management. Yet AI adoption often begins somewhere else: technology, operations, digital, or individual business units. That creates a risk.

A company can successfully deploy AI and still be completely unprepared for its impact on the workforce. HR needs to be in the room earlier. Not only when roles are eliminated or employees need training, but also when leaders decide how to divide work between people and machines.

The most forward-thinking organizations may eventually treat AI workforce design as a shared responsibility across the CEO, CHRO, CIO, and CX leadership. Because no single function can see the whole picture.

The Training Paradox

An irony lies at the heart of AI adoption. Companies want experienced employees with strong judgment. Then they automate the routine work through which less-experienced employees traditionally developed that judgment. That creates a potential training gap. If AI handles the easy cases, how does a new employee learn to recognize an unusual one? If AI writes the first draft, how does someone develop the ability to write a great one? If AI diagnoses the problem, how does an employee learn to diagnose the diagnosis?

These questions matter because expertise is rarely acquired through classroom training alone. It is built through exposure, repetition, mistakes, coaching, and increasingly difficult situations. Workforce reinvention therefore cannot simply mean sending employees to AI training. It requires leaders to rethink how people learn the work.

Organizations may need new apprenticeship models, simulation environments, mentoring programs, rotational assignments, and deliberately preserved opportunities for employees to practice judgment. The goal should not be to make people dependent on AI. It should be to make people better because they know how to work with it.

What Customers Will Expect from the New Workforce

CX leaders cannot ignore another side of this equation: the customer. Consumers are becoming more comfortable with AI, but comfort does not mean unconditional acceptance. Gartner’s 2026 research found that 50% of customers say interactions are easier when companies use GenAI, yet 87% say companies using GenAI for customer service must provide an option to reach a human agent.⁹

That is not a contradiction. It is a design principle. Customers are perfectly willing to let AI handle something when AI is useful. They become less willing when AI becomes an obstacle.

The same principle applies to the workforce. Employees will likely embrace AI when it removes pointless work, gives them better information, helps them resolve problems, and creates room for more meaningful responsibilities. They will resist it when it simply increases monitoring, removes autonomy, or makes them accountable for decisions they cannot control. In both cases, the issue is not whether AI is involved. It is whether the human understands where they stand in the system.

The More We Build with AI, the More We Have to Build Around People

AI adoption can create enormous capacity. But capacity is not the same thing as value. The question is what an organization does with the capacity it creates. Does it simply reduce headcount? Does it absorb more customer volume? Does it invest in better experiences and give employees more time to solve the problems that automation cannot? Does it create entirely new roles? Or does it do some combination of all of them?

The answer will be different for every organization. But leaders will increasingly need to make the answer intentionally. That means mapping the work, not just jobs. It means understanding which activities should be automated, which should be AI-assisted, and which should remain human-led. It means redesigning workforce plans around capabilities rather than historical staffing ratios.

It means reconsidering compensation, building new career paths, and measuring AI not only by productivity but by changes in employee capability, customer outcomes, and business performance. And it means giving employees a reason to believe reinvention is being done with them, not to them.

The Workforce Reinvention

So, What Are We Actually Building?

The AI conversation has spent years asking whether machines will replace people. That may be the wrong question. The more consequential question is whether organizations are willing to redesign themselves around what humans and machines can each do exceptionally well.

Once AI can answer the easy questions, execute routine tasks, and increasingly manage entire portions of a customer journey, the value of the human workforce will depend on what remains. Judgment, context, accountability, empathy, creativity, relationship-building, and problem-solving.

The ability to recognize that the answer on the screen is technically correct but completely wrong for the person standing in front of you. Those capabilities will matter. But they will not matter simply because they are human. They will matter if organizations deliberately build roles, cultures, training, career paths, and operating models that allow people to use them.

So, as AI takes more of the work, what kind of workforce are you choosing to build around it?

Article Links

1. https://www.gartner.com/en/newsroom/press-releases/2025-12-02-gartner-survey-finds-only-20-percent-of-customer-service-leaders-report-ai-driven-headcount-reduction?
2. https://www.gartner.com/en/newsroom/press-releases/2025-12-02-gartner-survey-finds-only-20-percent-of-customer-service-leaders-report-ai-driven-headcount-reduction?
3. https://newsroom.ibm.com/2026-09-21-new-ibm-chro-study-ai-puts-critical-thinking-at-the-center-of-workforce-priorities?
4. https://www.gartner.com/en/newsroom/press-releases/2026-08-04-gartner-survey-finds-87-percent-of-customers-say-companies-using-genai-for-customer-service-must-provide-access-to-a-human-agent0?
5. https://execsintheknow.com/knowledge-center/customer-experience-research/hot-topics-research/cx-agent-insights-perspectives-from-the-frontline/
6. https://execsintheknow.com/knowledge-center/customer-experience-research/hot-topics-research/cx-agent-insights-perspectives-from-the-frontline/
7. https://execsintheknow.com/knowledge-center/customer-experience-research/hot-topics-research/cx-agent-insights-perspectives-from-the-frontline/
8. https://www.ibm.com/thought-leadership/institute-business-value/en-us/c-suite-study/chro
9. https://www.gartner.com/en/newsroom/press-releases/2026-08-04-gartner-survey-finds-87-percent-of-customers-say-companies-using-genai-for-customer-service-must-provide-access-to-a-human-agent0