The rapid adoption of AI is changing how leadership is defined, assessed and developed

Nearly four years after the mainstream arrival of generative artificial intelligence, most organizations are still struggling to understand what it actually means for work. The dominant narrative has focused on job displacement: which roles will disappear, which will survive, and which new ones will emerge. This framing is intuitive, but incomplete. The real impact of AI is not primarily about eliminating jobs. It is about redefining them.

AI is transforming work at the level of tasks, not roles. It is unbundling jobs into their component parts, automating some, augmenting others, and leaving a smaller but more critical subset firmly in human hands. In doing so, it is quietly reshaping the nature of human contribution at work. The implication is not that humans are becoming obsolete, but that the bar for what constitutes valuable human input is rising.

This shift has profound consequences for leadership, talent and the broader labor market.

Many organizations are confusing activity with progress. According to survey findings presented by Nicholas Bloom of Stanford University, most surveyed firms reported using some form of AI technology, with overall adoption at about 69 percent across the sample. Yet these impressive figures often mask a more sobering reality: there is little evidence of corresponding productivity gains. Employees are frequently doing the same work as before, only faster or with less effort. Crucially, the time saved is rarely reinvested in higher-value activities.

This is not a productivity revolution. It is, at best, a more efficient version of the status quo. At worst, it is what one might call performative productivity, where the appearance of progress substitutes for actual value creation.

The underlying issue is that most organizations have not redesigned work. They have layered AI onto existing processes rather than rethinking those processes from first principles. As Peter Drucker famously observed, there is nothing so useless as doing efficiently what should not be done at all. AI risks amplifying precisely this problem, enabling organizations to do more of the wrong things, faster.

This is where leadership becomes critical.

If AI is the defining technological challenge of our time, then leadership is the defining organizational challenge. Leaders must decide not only how to deploy AI, but what work should exist in the first place. They must move beyond adoption metrics and focus on value creation. This requires a level of judgment that no algorithm can substitute for.

Paradoxically, as AI becomes more capable, the importance of human judgment increases. One reason is that AI generates what might be called artificial certainty. Large language models produce answers that are coherent, confident and often persuasive, even when they are incorrect. The risk is not that people will distrust AI, but that they will trust it too much.

Distinguishing between high-quality insight and plausible nonsense is fast becoming a core professional skill. It is also a leadership skill.

This raises an uncomfortable question. If the value of human work is shifting toward judgment, decision-making and problem framing, how well equipped are today’s leaders to meet that standard?

The evidence is not encouraging. For decades, leadership selection has been driven by flawed assumptions. Organizations tend to overvalue past performance, even though success is often context-specific and difficult to replicate. They place excessive emphasis on experience and technical expertise, despite strong evidence that personality and cognitive ability are more robust predictors of long-term potential. And they continue to reward confidence and charisma, traits that are easy to observe but only weakly related to effectiveness.

These biases help explain why so many organizations end up with leaders who look impressive but underperform. In an AI-enabled world, this problem is likely to become more visible, not less. When information and technical expertise are widely accessible, the marginal value of leadership shifts to areas where humans still have an advantage: setting direction, exercising judgment, building teams and influencing others.

This has important implications for how organizations identify and develop talent.

First, talent identification must become more data-driven. AI can play a useful role here, not as a replacement for human decision-making, but as a tool for improving it. By analyzing large datasets on leadership behavior and outcomes, organizations can identify patterns that are difficult to detect through intuition alone. This can help reduce bias and improve the accuracy of selection decisions.

However, data is only as useful as the framework used to interpret it. Without a clear understanding of what drives performance, more data can simply lead to more noise. This is where science matters. Decades of research in organizational psychology have identified a relatively small set of variables that reliably predict leadership effectiveness, including cognitive ability, personality traits and learning agility. Integrating these insights with AI-driven analytics offers a powerful way to improve talent decisions.

Second, leadership development needs to be rethought. Traditional approaches often focus on building strengths, under the assumption that maximizing what individuals are already good at will drive performance. In reality, many leadership failures are caused not by a lack of strengths, but by the presence of unaddressed weaknesses, particularly those linked to personality derailers. Traits such as overconfidence, volatility or excessive risk-taking can undermine effectiveness, especially under pressure.

In this context, the most effective development strategies are those that help leaders manage their risks, not just amplify their strengths. This is especially important in an AI-enabled environment, where the consequences of poor judgment can be magnified.

Third, organizations need to reconsider which capabilities are uniquely human and therefore worth investing in. As AI takes over routine cognitive tasks, the comparative advantage of humans lies in areas such as critical thinking, creativity, social influence and ethical reasoning. These are not new skills, but they are becoming more valuable.

Unfortunately, education systems and corporate training programs have been slow to adapt. They continue to emphasize knowledge over thinking skills, and technical proficiency over judgment. This mismatch risks producing a workforce that is not prepared for tomorrow’s challenges.

Finally, there is a broader societal dimension to consider. The transformation of work driven by AI raises important questions about governance, inequality, and social stability. Policymakers face the difficult task of balancing innovation with regulation, ensuring that the benefits of AI are widely shared while mitigating its risks.

This will require a more nuanced approach than the binary debates that often dominate public discourse. Overregulation can stifle innovation, while underregulation can lead to unintended consequences and negative ethical and societal outcomes. The goal should not be to control AI, but to shape its application in ways that enhance human capability and societal well-being.

In this sense, the challenge is not technological, but institutional. It is about how organizations and societies adapt to a new distribution of human and machine capabilities. The same is true at the business level. AI will not replace leaders but will expose them. It will make it easier to see who adds value and who does not. It will reduce the returns to superficial competence and increase the premium on genuine expertise and sound judgment.

For organizations willing to rethink how they define, assess and develop leadership, this represents an opportunity. For those that are not, it represents a risk.

The future of work will not be determined by AI alone. It will be determined by how effectively humans choose to use it. This is why effective leaders will be an integral part of the future.

The author is the chief science officer at Russell Reynolds Associates.

The author contributed this article to China Watch, a think tank powered by China Daily. The views do not necessarily reflect those of China Daily.

Contact the editor at editor@chinawatch.cn.