Leadership in the AI Era: Why Human Skills Matter More Than Ever

AI can write, code, analyze, and complete tasks at impressive speed. It can draft emails, create agendas, and prepare important meetings. But it still cannot do the hard work of leadership.

This is the central thesis of research published by McKinsey in January 2026, authored by Bob Sternfels (Global Managing Partner), Borge Brende (President of the World Economic Forum), and Daniel Pacthod (Senior Partner). It is a thesis that deserves attention.

The Leadership Paradox in the AI Era

We are living through an interesting paradox. The more AI advances in technical capabilities, the more important exclusively human skills become. Leadership remains deeply human — and more critical than ever, given the scope of change and uncertainty that organizations face today.

What AI still cannot do:

Set aspirations: Establish ambitious goals that mobilize the entire organization
Make tough calls: Make choices when values conflict and time is short
Build trust: Generate loyalty and engagement among diverse stakeholders
Generate genuinely new ideas: Create innovations that break existing patterns

These four elements are not just difficult to automate — they define what it means to lead.

The Central Shift: From Command to Context

McKinsey Global Institute research suggests that people, agents, and robots will increasingly work side by side. In this environment, traditional command-and-control approaches tend to fail.

The fundamental shift is this:

Traditional Model:

Leader as “the smartest person in the room”
Top-down, centralized decisions
Direct control over execution
Rigid information hierarchy

AI Era Model:

Leader as context creator and guardrail setter
Distributed decision rights
Clear definition of quality standards
Environment of trust and collaboration

This does not mean abandoning directive leadership. Research shows that organizations with strong leadership make decisions 60% faster in crisis moments. The secret is knowing when to apply each style.

Three Exclusively Human Domains

The research identifies three areas where only humans can lead effectively.

1. Setting Aspirations and Enrolling Others

Aspirations are exclusively human. A robot cannot set an ambitious goal for the entire organization.

Leaders can “read the room,” anticipate and interpret emotional reactions to change — something critical for mobilizing people and involving them in strategic plans. Only an empathetic leader can identify preferences and allocate the right people to the right projects.

Practical actions:

Articulate a vision that inspires and gives meaning to work
Translate strategy into goals that resonate with different teams
Use AI to draft messages, but personalize communication
Create alignment rituals that go beyond slides and dashboards
Celebrate milestones and recognize contributions individually

Leaders can use agents and machines to help draft messages, but they cannot delegate the definition of aspirations.

2. Demonstrating Judgment and Aligning Choices to Values

The ability to demonstrate good judgment is a distinctly human trait.

AI can summarize rules or outline risks, but its role is advisory, not authoritative. AI models do not carry real responsibility for their outputs. On the other hand, leaders in the physical world must be held accountable to employees, boards, investors, and partners.

When human judgment is irreplaceable:

Conflicting values: When organizational principles collide and there is no objectively “right” answer
Limited time: Decisions that need to be made without complete data or consensus
Important precedents: Choices that will define how the organization acts in future cases
Impact on people: Decisions that affect careers, well-being, or organizational culture

McKinsey research shows that leaders’ decisiveness, accountability, and demonstration of good judgment not only unlock trust and loyalty but are key predictors of the ability to create long-term value.

3. Designing for Non-Linear Outcomes

In a world where global, social, and technological forces change dynamically, the ability to foster new and creative ideas is a critical leadership muscle.

AI models are inference engines, optimized to generate the most likely next continuation of patterns they have already seen. This means they are excellent for incremental improvements — 20% better — but not for real transformation — 10x better.

Leaders must seek results that are 10 times better, not just 20% better.

The leader’s hard work:

Define an audacious brief that challenges the status quo
Establish clear guardrails without stifling creativity
Invite constructive dissent and listen to divergent voices
Hold the creative line when initial iterations are confusing
Recognize when AI outputs will lead to real breakthroughs vs. incremental ones

The antithesis of “playing not to lose”: leaders must continually revise, revise, and create new architectures, narratives, and conditions for their organizations.

Four Imperatives for Developing Leaders

How can organizations develop leaders prepared for the AI era?

Imperative 1: Know What Attributes You Are Looking For

Make explicit the leadership attributes your company needs now and the behaviors you will reward.

If economic and competitive shocks are particularly frequent or acute in your sector, you may want to focus on resilience and optimism as key characteristics.

The evaluation system must change: more audition than interview — live scenarios with incomplete information, structured questions that test values-based judgment, and rapid moves to challenging roles that reveal trajectory.

Critical attributes by context:

Context	Priority Attributes
High Volatility	Resilience, adaptability, rapid decision-making
Rapid Growth	Scalability of thinking, effective delegation
Digital Transformation	Technical fluency, change management, human-AI collaboration
Culture in Crisis	Integrity, transparent communication, empathy

Imperative 2: Create a Step-Change in Learning Culture

Establish a culture where premortems, post-action reviews, and other feedback mechanisms are the norm, not the exception.

This approach has long been standard in the software industry and medical community: achievements are celebrated, failures are reviewed meticulously, and lessons are codified.

Brad Smith, former CEO of Intuit, describes it this way: “You skip levels and go straight to the front line of the area you are trying to learn. Cut everyone else out and eliminate the filter.”

Practices of effective leaders:

Schedule regular meetings with employees several levels below
Ask: “What is improving? What is getting worse? What are you afraid to tell me?”
Create forums where senior leaders engage directly with high-potential talent
Share management challenges and crowdsource solutions
Transform failures into documented case studies

The model: learn a little, test a little, learn a lot. Rapid experimentation cycles with structured feedback loops.

Imperative 3: Invest in Trust and Servant Leadership

Organizations must actively cultivate core leadership qualities such as wisdom, empathy, and trust — giving these attributes the same attention they give to new IT systems or operating models.

This means providing time for leaders to do the inner work necessary to lead others effectively:

Reflect on purpose and meaning
Share insights with other C-level leaders
Consider what success means for themselves and for the organization

Satya Nadella, CEO of Microsoft, cites a lesson from his father: “If the next CEO of Microsoft can be more successful than me, then maybe I have done my job right.”

Signs of servant leadership:

Publicly celebrate or promote leaders who demonstrate commitment to the broader mission
Prioritize successor development over personal achievements
Create meaningful transitions and active mentorships
Recognize that the leadership journey is as important as the results

Imperative 4: Protect Your Time and Energy for Sustained Performance

The best-performing leaders create conditions that allow them to reach their personal best at peak moments.

They recognize that, throughout their leadership tenures, some moments are simply more important than others — so they optimize for these critical inflection points.

A global technology CEO keeps 20% of their calendar free to respond adequately to situations as they arise.

Energy protection practices:

Fiercely protect the agenda to focus on tasks only you can do
Explicitly reserve time for recovery and regeneration
Identify your critical inflection points and prepare for them
Delegate decisions that others can make as well as you
Create systems that filter information without losing important signals

Like elite athletes, leaders must manage energy, not just time. Periods of maximum intensity require periods of adequate recovery.

Identifying High-Potential Talent

Aspiration, judgment, and creativity are “human-only” leadership traits — characteristics that can provide an irreplaceable competitive advantage, especially when amplified by AI.

Organizations must actively identify and develop individuals who demonstrate intrinsic characteristics such as:

Resilience in the face of adversity
Willingness to learn from mistakes
Ability to work in teams that will include humans and AI agents

McKinsey research shows that these characteristics are better predictors of long-term success than formal credentials.

Breaking the “Paper Ceiling”

Many organizations block talent due to lack of formal credentials. This requires:

Looking beyond diplomas and certifications when evaluating candidates
Emphasizing skills-based hiring
Valuing real and relevant experience
Prioritizing intrinsic characteristics that cross functions and technologies

Harvard and Burning Glass research shows that 85% of companies say they do skills-based hiring, but only 0.14% of hires are actually affected. Execution matters more than intention.

A Balanced Perspective

Every analysis deserves critical scrutiny. Consider these additional perspectives:

What the research does not address:

Economic pressures may force AI choices over human leadership
The boundary between human and AI capabilities continues to move
Different sectors require different combinations of styles
Implementation is harder than the recommendations suggest

As economist Andrew J. Scott observes: “As machines get better at being machines, humans need to get better at being humans.”

Questions for Your Organization

Before closing, some questions for reflection:

Which decisions in our organization require irreplaceable human judgment?
Where are we using AI to think with us vs. think for us?
How are we developing the next generation of leaders to collaborate with AI?
Do our evaluation systems identify the right characteristics?
Are we investing in the “inner work” of our leaders?

Conclusion

AI can transform how we work, but only human leaders can determine why we work and what we are trying to achieve.

The three domains — aspirations, judgment, and non-linear creativity — are not just areas where humans are still better. They are areas where the nature of the work demands qualities that machines, by definition, cannot have: moral responsibility, genuine empathy, and the ability to inspire other human beings.

The four imperatives — define attributes, transform learning, invest in trust, and protect energy — are not new. They are leadership principles that have stood the test of time. What changes is the urgency of applying them.

In the AI era, leadership does not become less important. It becomes the differentiator.

At Victorino Group, we help leaders navigate digital transformation with strategic clarity. If you want to implement AI with governance in your team, let’s talk.