TECH & INNOVATION

By FedEx | November 6, 2025

Successful AI leadership requires far more than investing in the right technology. Kriti Kapoor, a digital innovation and workforce transformation strategist and author of “The Mid-Career Digital Shift”, shares her insights on how leaders can navigate change, build AI-ready teams, and lead with greater confidence in the age of AI.

As AI adoption accelerates across virtually every sector, businesses of all sizes are racing to transform their workflows. But beyond the hype, leaders face a million-dollar question: Do AI productivity tools actually translate into meaningful business value?

Few people are more qualified to help answer this question than Kriti Kapoor. Throughout a nearly 30-year-long career in enterprise technology – spanning executive-level roles at Dell, HP, and Microsoft – Kriti has upskilled over 250,000 professionals and students. During her time leading the community team of HP’s Intelligent Automation Center of Excellence, she drove large-scale adoption of digital innovation and workflow automation, ultimately leading to more than 2,000 automations and over 250,000 hours saved each year across the organization.

Today, she’s bringing this wealth of experience to help leaders bridge the gap between AI investment and business impact. Kriti helps enterprises find where digital innovation pays off, equip leaders to guide teams through change, and design human-AI collaboration. Recently, she also shared her expertise with young entrepreneurs at the 2026 FedEx/JA International Trade Challenge Asia Pacific finals, where she joined the judging panel to evaluate student-led business innovations.

Here, Kriti offers her insights on how leaders can unlock tangible value from AI tools, navigate workforce transformation effectively, and build the leadership skills and capabilities needed for an AI-driven world.

What does it mean to be a good leader in this age of AI? Is it different from what it meant five to 10 years ago?

Kriti Kapoor: Yes, although the fundamentals of good leadership remain the same. People still need leaders they can trust, who take responsibility, provide direction, and help them grow. What’s changed is the pace of change, the complexity of decisions, and the impact those decisions have on how people work.

In “The Mid-Career Digital Shift”, I explore three key dimensions of leadership: Navigate, Operate, and Lead. Put simply, leaders need to know where they're going, work out how to get there, and help people move forward together.

First, navigating change requires greater clarity and judgment. AI can generate dozens of ideas and possibilities in minutes, but leaders still need to determine which opportunities are worth pursuing and why.

Second, operating effectively requires continuous learning and adaptability. Five or 10 years ago, leaders could often rely on established processes and expertise for longer periods. Today, they need to experiment, learn alongside their teams, and rethink how work gets done. That means developing their own AI fluency, understanding the technology's capabilities and limitations, and being comfortable admitting, "I'm still figuring this out, too."

Third, leading people through change requires empathy, trust, and honest communication. This has always been central to leadership, but AI makes it more urgent. As roles evolve, employees naturally ask: "Where do I fit? Is my experience still valuable? What will I need to learn?"

These are legitimate concerns that cannot be addressed simply by encouraging people to "embrace AI." Leaders need to have honest conversations about what is changing, acknowledge uncertainty, involve employees in shaping new ways of working, and provide meaningful opportunities to develop new capabilities.

Are there leadership skills that are becoming obsolete because of AI?

I'd hesitate to call them obsolete, but some traditional sources of leadership authority are certainly losing their value. Being the person with all the answers, controlling access to information, or overseeing routine reporting used to be a source of influence. AI is making information and analytical capabilities much more accessible. Leaders still need to understand the work, but their value increasingly comes from what they do with that information.

I see three capabilities becoming particularly important: judgment, digital fluency, and the ability to bring people together.

First, judgment. AI can produce remarkably convincing answers, but leaders remain responsible for deciding whether those answers make sense. The ability to ask better questions, challenge assumptions, and make sound decisions becomes more valuable as AI takes on more analytical work.

Second, digital fluency. Leaders don't need to master every AI tool or become technical experts, but they do need to understand what the technology can do, where it falls short, and how to apply it responsibly.

Leaders will increasingly need to decide which tasks to delegate to AI, where human oversight is essential, how to evaluate the quality of AI-generated work, and who remains accountable when something goes wrong. AI safety and governance are part of that responsibility. Leaders need to establish clear boundaries around data privacy, decision-making authority, and human oversight.

Third, the ability to bring people together. Today's workforce spans multiple generations, with different experiences, expectations, and levels of confidence in technology. We need to move beyond assumptions that younger employees are naturally AI-savvy or that older employees are resistant to change. Good leaders create opportunities for people to learn from one another, combine their strengths, and build confidence together.

This is also where experience becomes a competitive advantage, provided we continue updating it. An experienced operations leader may understand why a process exists, where exceptions occur, and what could go wrong. But they must also be willing to challenge established ways of working and explore better alternatives.

What is the biggest misconception executives have about leading an AI transformation?

The biggest misconception is that investing in AI technology and getting it up and running means you've achieved transformation.

I see AI transformation as fundamentally a business and workforce transformation. You're changing how work gets done, how decisions are made, and what people need to succeed. That requires sustained leadership, organizational change, and capability building long after the technology goes live.

Buying AI productivity tools and launching pilots are useful starting points. But the real question is: What is actually improving, and for whom? Across the digital transformation programs I've led in Asia, Europe, and the US, collectively delivering more than USD 80 million in value, I've seen the same lesson repeatedly: technology, business processes, and people's capabilities need to evolve together.

In “The Mid-Career Digital Shift”, I introduce a Work-to-Impact Journey that encourages leaders to examine how work actually flows, identify friction and opportunities, redesign processes with AI and human capabilities in mind, and measure the outcomes that matter.

For executives, this means staying actively involved beyond approving the investment. Give employees time to experiment and develop new capabilities. Remove barriers between teams, and ensure that performance measures and incentives reinforce the changes you're asking people to make.

Crucially, involve the people doing the work from the beginning. They understand the exceptions, workarounds, customer frustrations, and operational realities that rarely appear on a process map. Without their input, it's easy to design a process that looks impressive on paper but fails in practice.

There's also a broader opportunity that executives sometimes overlook. Much of the AI conversation focuses on doing existing work faster or cheaper. That's important, but leaders should also ask: What can we now do that wasn't possible before? Perhaps it's delivering a more personalized customer experience, creating an entirely new service, or experimenting with ideas that previously required too much time or investment.

Where are companies seeing real productivity gains from AI, versus gains that are mostly anecdotal?

The clearest evidence of AI productivity gains comes from specific, measurable tasks where we can compare performance with and without AI. I'm more cautious about broad claims of productivity gains across an entire organization. When someone says, "Our employees are saving several hours a week," my next question is: What happened to those hours? Did customers receive faster service? Did quality improve? Could employees handle more complex work?

Consider customer support. A study published in The Quarterly Journal of Economics, covering more than 5,000 support agents, found that AI assistance increased the number of issues resolved per hour by 15% on average. Less experienced employees benefited the most, suggesting that AI can also help people develop their capabilities faster.

In knowledge work, the picture is more nuanced. A Harvard Business School and Boston Consulting Group study found that consultants using AI completed tasks within the technology's capabilities more than 25% faster, with improvements in the quality of their work. However, when working on a task beyond AI's capabilities, those using AI were more likely to reach incorrect conclusions. Knowing when to use AI, and when to question it, is just as important as knowing how to use it.

Software development offers another useful caution. In a 2025 study by METR, experienced developers working on familiar codebases took 19% longer to complete tasks when using AI tools, even though they believed the tools had made them faster. The findings were specific to that study and its tools, but they highlight why perceived productivity and actual performance.

We also need to account for the time spent reviewing, correcting, and integrating AI-generated work, along with the cost of the technology itself. Generating a report in five minutes sounds impressive, but the real measure is how long it takes to produce an accurate, usable result.

For businesses looking to improve productivity with AI, my advice is straightforward: start with a clearly defined problem, establish a baseline, test AI against it, and measure the results. Look at speed, quality, cost, and the impact on customers and employees. If the results are consistently positive, build from there.

How can companies ensure employees are adopting AI in a meaningful way, instead of doing it for the sake of it?

Start by giving people a clear, practical reason to use AI. If the instruction is simply "Use AI more," you'll get activity, but you may not get better outcomes. Instead, asking employees to "Help us prepare accurate customer proposals faster" gives them a specific problem to solve while reinforcing that quality still matters.

I'd also look beyond the number of AI licenses purchased, login frequency, or usage statistics. These are useful indicators of access and engagement, but they tell us very little about whether AI is improving how people work.

To understand whether AI adoption is creating meaningful value, I'd ask four questions:

1. Can employees recognize where AI is useful, apply it appropriately, and critically evaluate its output?
2. Are people actually changing how work gets done, or simply adding AI to existing processes?
3. Are we seeing measurable improvements in speed, quality, cost, customer experience, or employee effectiveness?
4. What are we doing with the time and capacity AI frees up?

That last question is especially important. Perhaps the team could spend more time with customers, tackle complex problems, develop new skills, or find relief from an unsustainable workload. Leaders need to be transparent about how productivity gains will be used. If every efficiency improvement simply results in more work, employees may become reluctant to share what they've achieved.

During my time leading the global automation community at HP, I helped grow our citizen developer community from about 250 to over 5,000 employees. What made this meaningful was that employees were empowered to solve problems they understood firsthand, building automations that improved processes and freed up valuable time. We focused on the work being improved, the problems being solved, and the value being created.

I'd apply the same principles to AI workforce transformation. Start with a real business problem, give employees the tools, time, and support to experiment, and create opportunities to learn from the results. This kind of practical, continuous learning builds both AI fluency and confidence. It also helps organizations identify successful applications that can be shared and scaled across teams.

Finally, leaders need to create an environment where employees feel comfortable saying, "AI introduced errors," or "This task works better without it." That feedback is essential to responsible adoption and continuous improvement.

How can AI tools help leaders become more effective in their roles?

AI can help leaders make better-informed decisions, work more effectively, and free up time for the responsibilities that need their attention most. In “The Mid-Career Digital Shift”, I describe four ways of working with AI: AI can assist, advise, execute tasks, or serve as a thinking partner to help us explore ideas and challenge our assumptions.

For example, before making an important decision, leaders can use AI to compare options, examine different scenarios, identify gaps in their thinking, and test assumptions. Before a difficult conversation, they can rehearse their approach and explore how the other person might respond. After a meeting, AI can help summarize discussions, capture decisions, and draft follow-up actions.

However, effective AI leadership requires knowing what to delegate, what to verify, and what must remain a human responsibility. The degree of autonomy we give AI should reflect the task, its risks, and the consequences of getting it wrong. Decisions involving significant business risk, ethics, employee well-being, or strategic direction require appropriate human judgment and accountability.

The growing use of AI agents also brings new leadership responsibilities. Managers need to ensure that AI-generated work is reliable, establish clear boundaries and oversight, and prevent over-reliance on technology from weakening employees' skills or reducing meaningful human interaction.

AI can also support leaders in developing their people. For instance, a manager preparing for a career development conversation might use AI to generate thoughtful coaching questions, identify relevant learning opportunities, or suggest ways an employee could practice new skills through their work. But those suggestions still need to be adapted to the individual's strengths, aspirations, and circumstances.

Ultimately, the greatest opportunity is to use AI to become a more effective and more human-centered leader. My advice is to reinvest some of the time AI frees up in listening to employees, coaching teams, strengthening customer relationships, and thinking more strategically.

How can leaders develop and retain wisdom in an age where AI is rapidly taking over analytical tasks?

For me, wisdom develops through experience, making decisions, seeing the consequences, and reflecting honestly on what we got right or wrong. AI can help us analyze information, explore options, and identify patterns, but we still need to exercise our own judgment.

One risk is that a confident, well-written AI recommendation can appear more reliable than it actually is. Leaders need to question the assumptions behind it, recognize what might be missing, and consider the broader consequences.

Imagine AI recommends closing a customer service channel because it is expensive to operate and relatively few customers use it. The analysis might be financially sound, but a wise leader would ask: Who are those customers? Do they have accessible alternatives? How would closing the channel affect their experience and trust in the company? The numbers matter, but they rarely tell the whole story.

I'd encourage leaders to develop three simple practices.

• Form your own perspective before consulting AI. Then use AI to challenge your reasoning, explore alternatives, and identify what you may have overlooked. This helps strengthen independent thinking rather than becoming overly dependent on AI-generated recommendations.
• Stay close to employees, customers, and partners. Spend time with the people who experience the consequences of your decisions. Listen to their concerns, understand their circumstances, and observe what happens in practice. Some of the most valuable insights never appear in a dashboard or management report.
• Keep a simple decision journal. Record the decision you made, the reasoning behind it, and what you expected to happen. Revisit it later to compare expectations with actual outcomes. A good outcome doesn't always mean you made a good decision. Equally, a well-reasoned decision can produce a disappointing result. Learning to distinguish between decision quality and outcomes helps improve judgment over time.