The 3 trials of leadership in the age of AI

We know that AI is reshaping work. In a 2025 interview with CNBC, Salesforce CEO Marc Benioff stated that AI is now doing 30%–50% of the work in areas like engineering, coding and support at the company. The reality is that AI will affect work, but its efficacy is a completely different question. A Boston Consulting Group study revealed that 74% of companies are failing to extract meaningful value from AI after two years.

In a 2025 study, McKinsey notes that “the challenge of AI in the workplace is not a technology challenge—it is a business challenge,” requiring companies to align leadership, redesign workflows and rewire their organizations for change. The realities of AI present three leadership trials for organizations; unlike previous shifts that required mere recalibrations of leadership capabilities, these are more existential. They require us to reshape how we conceptualize leadership.

In Ancient Greek mythology, the Labors of Hercules were designed to test him: his strength, his ingenuity and his moral discipline. AI presents Herculean-styled trials for leaders today, and like Hercules, the leaders who make it through will need to innovate new methods, collaborate with others and outthink the problem. AI will not just reshape work; it will reshape leadership. Beyond the immediate pressures for adoption and enablement of AI fluency, the best organizations will need to contend with these three leadership trials to thrive.

The three leadership trials surround leadership identity, leadership techniques and leadership governance. Where Hercules had to defeat monsters with more than brute force, the Leadership Trial of Identity will challenge the traditional core of what organizations have valued in leaders, and how they have selected, promoted and developed them. Many of Hercules’ tasks required ingenuity. We know that AI will require leaders to rethink work and their place in it radically; this is the Trial of Leadership Technique—where leaders need to rethink workflows and adapt to broader spans of control, and leading humans and AI, together. Many of Hercules’ trials were imposed by Fate; the final Leadership Trial surrounds governance. AI presents material governance, structural and moral dilemmas that today’s Boards are ill-equipped to address.

The Leadership Trial of Identity: Leadership requires different skills and different people

For decades, organizational focus on leadership has been firmly divided into two camps of competence: The “hard skills” of business knowledge, strategic thinking and analytic prowess and the “soft skills” that surround working with others—collaboration, coaching and enabling teams and creative thinking. Historically, the “hard skills” have prevailed as the primary competencies that organizations use to promote leaders. “Soft skills” were acknowledged but often pushed to the back of the queue.

The acceleration of AI in the workplace is about to upend decades of focus on what constitutes leadership excellence. Many of the hard skills that have been fundamental to how we measure performance and develop leaders are about to become over-the-counter commodities as a result of AI. Analytics, strategic assessment and business review are elemental to AI. Interestingly, the soft skills that have been grossly underinvested in are about to become the key differentiators in leadership excellence, and organizations are woefully underprepared for this trial.

A recent study by MIT looked at the limitations of AI as a means to understand what will be required of humans in the future; the “un-AI-able” skills. Guess what? The “hard skills” that have been fundamental to how we have valued and assessed leaders are about to become severely discounted. The EPOCH study identified five human capabilities that will become more critical in the future, primarily because AI can’t replace them:

• Empathy and emotional intelligence
• Presence, networking and connectedness
• Opinion, judgment and ethics
• Creativity and imagination
• Hope, vision and leadership.

Most AI-forward organizations are investing heavily in identifying new leadership capabilities. Salesforce has implemented the 10 Enterprise Skills as a strategic initiative, focusing on reskilling its workforce in these key competencies. The three domains they are targeting are “human skills”, “agent skills” and “business skills.” The human skills that Salesforce is focusing on include:

• adaptability
• accountability
• collaboration
• emotional intelligence

While technology can do a lot in the agentic AI enterprise, it cannot replicate the ability to build trust, inspire a team, creatively adapt or foster psychological safety. Organizations must face the reality that they have grossly undervalued these skills, and may have an entire talent ecosystem that values the wrong leadership capabilities.

Microsoft’s Future of Work Team has unpacked much of the nuance of how they see work shifting. Their consideration of team leaders, however, is particularly striking; with many on the team questioning if managers will be needed in the future at all. Their ultimate conclusion is that if anyone remains in a “management” role, their primary (if not lone) capability will surround coaching and developing others.

Consider how different that reality is from what we see from managers today; this is the core of the Leadership Trial of Identity—leaders will need to fundamentally be different people with entirely different skill sets.

Leaders’ ability to create connection, community and to help people grow is emerging in other quarters, too. In a recent body of research, the Institute for Corporate Productivity (i4cp)—the global HR research organization—identified the ability to create a powerful team culture as the single greatest driver of high performance.

Leaders who could connect with and build social norms and strong cultures of accountability in their teams massively outperformed task managers and the hyper-demanding bosses. Leadership requires social skills and the ability to create greatness in others, not merely technical dominance.

The first leadership trial will challenge the identity of leaders. Organizations may have to face the reality that their leaders are ill-equipped for the task ahead and that they have developed and promoted people on capability sets that are no longer relevant.

Questions to understand the shape of the Trials

The Leadership Trial of Identity

• How do you redesign leadership competency frameworks and succession criteria to reprioritize soft skills right now?
• How do your leadership assessment approaches and talent systems need to be refined to identify evolving competencies (community building, unlocking energy, ability to develop others)?
• How do you identify and remove leaders who are stylistically mismatched for future requirements, and may impede the evolution and transition in leadership styles of others?

The Leadership Trial of Technique

• How do you reconsider work to identify critical work products, rethink how to deliver them and the “shape” of your organization with combined human and AI teams?
• How do you clarify the most essential leadership requirements with increasing spans of control?
• How do you simultaneously enable innovation and manage risks at an operational level, so that scaled and accelerated processes don’t bring down the entire enterprise?

The Leadership Trial of Governance

• What are the critical agenda items that the board needs to focus on to stay balanced?
• How do you rethink board composition and its governance requirements amidst risks and shifts in AI?
• How do executives and boards collaborate most effectively to manage and respond to increasing uncertainty and requirements, or faster and more focused oversight?

The Leadership Trial of Technique: Leading a blended AI and human workforce

While the first “Trial” of leadership is deeply human, the second requires the leadership techniques to manage a blended workforce. The tasks that were historically managed by leaders in organizations don’t just disappear; they get reshaped, automated and are driven by algorithms buried in Agentic AI. Spans of control are likely to expand, and leaders are going to be asked to master the technique to understand work in ways never imagined, as they oversee the activities of larger and larger teams comprised of both AI agents and humans. Very few leaders have the technical skill and know-how to understand the risks and drivers of this world, and fewer still know how to “manage” these blended teams.

In discussions with C-suite executives, the shape of the workforce and organizational design is a material question. One global law firm is postulating that their workforce will ultimately look like a capital “I,” with a larger, more senior layer, a skinny middle and a larger pool of junior talent. Other organizations are considering a more diamond-shaped structure, as they foresee the challenges of oversight presented by mixed teams. The logic of these shifts is built around very distinctive roles for organizational layers, which will be considerably different.

The requirements of each layer, and their organizational value, are shifting. Junior talent (graduates, for instance) tends to be AI-native, as they use AI as a feature of their lives; they know how to interact with it, how to guide it and how to get the best from it.

Middle management, where they exist, will need to have a much stronger presence and influence in the organization. They will lead larger, flatter teams and must have the ability to understand AI capabilities while matching them with a human workforce around a unified vision.

Senior leaders must have the judgment to guide some of the most material questions that organizations will need to face. How these three worlds come together, and the skills to translate this multitude of shifts into a coherent whole, is central to the Trial of Leadership Technique.

At its most fundamental level, this particular Trial of AI will ask leaders to answer one critical question: “What is work?” While organizations scramble to create AI fluency in the first instance, that focus is insufficient. Leaders will be asked to oversee performance in different ways, rethink “work” and scope projects at a level never seen and learn to manage an entirely new world of risks. They will need to cultivate techniques and approaches that allow them to observe and manage a new world, including things like:

• Redefining performance: Leaders will need to fundamentally rethink how they define performance. How a leader assesses the performance of agentic AI and humans working together will look different. For instance, a traditional metric like “Time to Resolve” is meaningless in an AI-enabled workflow. The management of AI will require different lenses and metrics that assess things such as credibility, bias and governance. While humans managing more complex work will no longer be measured on how quickly they turn queries around, new metrics will emerge around innovative problem solving and identification, and how effectively they work with other humans.
• Rethinking “work”: Leaders will need to develop a level of AI fluency that allows them to identify opportunities to redesign work locally. They will need to deeply understand the capabilities and limitations of AI. Assigning tasks to AI requires that work be deconstructed and put together in a way that may look very different to its current state. It requires leaders to have keen oversight and an awareness of markers when things go awry and agile enough to pivot quickly as new opportunities and new patterns come to the fore.
• Rethinking “teams”: Leaders will need to rethink how they design their teams. The impact of blended teams of humans and AI working together requires knowing when and how to allocate tasks to AI and when to have a human in the loop. Further, capacity planning must shift from an annual discussion around headcount requirements, to a much more nimble, fast-moving, and arguably ‘local” discussion of “cost to serve” with a heavier reliance on a contingent workforce. Microsoft’s Future of Work Team has an even more complicated view of this problem. Not only will teams be blended AI and human, but they will be constantly in flux. The workforce will become multi-modal—gig workers, AI agents, engineers, architects, consultants, process experts will need to be integrated into the team and leveraged as workers; and teams will form, disband and reform with increasing speed. The ability to create teams and enable them quickly and effectively will be fundamental to this Trial.
• Risk management: AI presents a major risk for leaders, as work products can be done instantaneously and at incredibly high volume, while trust can be eroded in an instant. Ensuring AI operates without bias, within guardrails and maintaining privacy is critical. Leaders need to be actively addressing the risks of AI, and they also have to be dialed into the bigger question: Who is ultimately accountable if AI makes a mistake?

The second Trial of Leadership requires a rethinking of the techniques that leaders use at an operational level; everything from metrics and task assignment to team construction and risk mitigation.

The Trial of Governance: Reshaping board leadership

The leadership of boards is not frequently considered as a “leadership trial” within organizations. Boards sit on high and approve and review long-term plans with executive teams in the interests of shareholder value; this has been the way for decades. Like Hercules, however, how executives lead boards is a Trial that AI will present. Most boards are disproportionately populated by accountants and bankers; the expertise that they have historically brought to capital matters have proven valuable, but they risk being too captive to a world that doesn’t exist or being too eager to do things that they don’t understand. How executives lead their boards, inverting the traditional hierarchy, will be a critical trial for many organizations.

In the rapidly shifting world of AI, the leadership of boards, and how CEOs and other senior executives focus boards on the most critical drivers of performance becomes vitally important. While many boards are seeking to accelerate their “tech savviness” and “AI fluency”, this is woefully inadequate and doesn’t capture the governance challenges presented by the first two Trials. Only 17% of corporate directors reported receiving training on GenAI, with 93% of surveyed corporate directors saying they lack confidence in their board peers’ comfort with using generative AI.

Boards are being asked to govern and provide long-term thinking on something that they barely understand; the support and leadership required around them is vital. Further, organizations that tackle the first two Leadership Trials need to recognize that their boards (and their governance structures) don’t adequately address the new realities. Boards will need to rethink oversight of the operational risks in an AI world; they will need to have a depth of insight into leadership capability and team dynamics that they have historically undervalued. Most critically, board membership may be woefully mispopulated to manage the risks of an AI-enabled organization.

Boards need to be the ballast of long-term thinking and organizational health, yet overly reactive stances are emerging in board leadership around AI. i4cp undertook a piece of research with corporate board members asking them to identify the factors that would give them the greatest confidence that their company was “future ready.” The results reveal a profound imbalance in how boards are thinking about “future ready”. The top four responses, by an overwhelming margin, surrounded solely AI. While it is wonderful to see board members responsive to trends in markets, when boards become reactive, it is dangerous.

Board leadership that is fixated on AI, and not systemic and balanced in their view of enterprise health, represent a critical risk to shareholder value. As an illustration, the bottom-rated items on this same survey surrounded succession and critical role pipeline, metrics on cultural adaptability and change readiness, and mapping of enterprise skills and subject matter expertise. All of the research points to the reality that the bottom rated priorities for boards are the most important for successful AI transformations and long-term organizational health, yet boards are severely discounting them, at scale.

Why does this matter? According to the World Economic Forum’s 2025 Future of Jobs Report, the two biggest barriers to organizational transformation are skills gaps in the marketplace and culture/change readiness. The Trial of Board Leadership puts a spotlight on the need for a holistic and systemic view from boards, not overreaction to a single trend or buzz word. We know from i4cp research that “future-ready” organizations require three things: cultural readiness, AI readiness and skills readiness. Boards that over-fixate solely on AI are a risk to their organizations.

Anxious boards that lack fluency and familiarity with AI require guidance and leadership from their executives. According to research, over 90% of corporate directors lack a high degree of confidence that corporate leadership has articulated a clear vision for the company’s future with AI, and that the company has programs in place to ensure the skills needed for the next two years. Boards need leadership from their executives that helps them understand what is important and how to look at it, or organizations need to go about radically resetting who is on their boards, and how they govern.

Organizational leaders need to push boards to maintain a balanced view and not become fixated merely on AI. As an illustration, a study of chief technology officers revealed that 93% see the barrier to data and AI adoption as cultural, not technical. A broad-based board obsession with culture might be a better focal point than AI. The final trial of leadership in the era of AI will require senior leaders to lead upwards and shape thinking in boards that considers the holistic system of performance, against the backdrop of AI and shifting approaches to risk.

Leadership is as much on trial from AI as any other part of business

While AI is being touted and feared as a radical disruptor that will challenge our very conceptualization of work, it presents something equally existential to our conceptualization of leadership. Like Hercules and his Labors, executives today face some critical trials. The three trials represent a generational threat to how we conceptualize what good leadership looks like, how we focus leaders on what is most important and how we help boards effectively balance their reactions and develop new approaches to governance.

In Greek mythology, the trials transform the hero. Hercules emerges as a different person from his Labors. AI presents significant trials for organizations, executives and their leaders. How effectively they clarify and address the trials will determine the shape and effectiveness of their transformation. The organizations and leaders who succeed will require their leaders at multiple levels to reimagine what “leadership” is, and emerge transformed.