By Vedika Lal, Zuzanna Staniszewska, and Géraldine Galindo
Artificial intelligence is transcending its original role as a tool designed to support humans and is increasingly performing leadership functions. What skills and competencies, then, will be demanded of human leaders in this brave new world?
“Our future is one where AI assumes leadership roles.” When Van Quaquebeke and Gerpott, pioneering researchers in leadership, wrote this in 2023, it sounded like provocation. Two years later, it sounds more like description. Across industries, AI no longer just supports leaders, it acts like one.
Rapid technological progress and advances in artificial intelligence and big data are supporting humans in analysing and interpreting large volumes of complex data, such as medical or corporate performance data1. As such, these technologies are transcending their original role as tools designed to support humans in everyday tasks, both inside and outside the workplace, and are increasingly assuming more intricate roles in which they can perform leadership functions.
AI has begun to replace and displace humans across a range of standardised managerial functions.
More specifically, AI has begun, and is likely to continue, to replace and displace humans across a range of standardised managerial functions , such as allocating tasks and resources, planning shifts, appraising performance, analysing team dynamics, determining compensation, and even making selection, promotion, and retention decisions. These changes in jobs and skills are leading to a rethinking of how they are managed.
Different forms of leadership
Leadership issues are at the forefront of considering these changes, but also of driving them forward, and this is evident by increasing AI integration beginning to demonstrate various forms of leadership.
Van Quaquebeke and Gerpott offer a useful distinction: AI can support, augment, or even substitute for human leadership.
Already, it is doing all three:
- Through task allocation and real-time feedback or guidance, AI performs task-oriented leadership
- Through sentiment monitoring and behavioural nudging, it enacts relationship-oriented leadership
- Through the generation of strategic narratives, persuasive messages, and personalised communication, it performs change-oriented leadership3.
Thus, while the leader might be human, the logic shaping those decisions is becoming increasingly algorithmic. This suggests that organisational values are increasingly being embedded in algorithms, and that AI has become capable of performing functions once considered resistant to automation4.
What, then, is left for human leaders?
With several leadership functions increasingly being taken over by AI, it is highly likely that organisations will require fewer human leaders, particularly at lower and middle management levels.
This does not imply that human leaders will no longer be needed entirely. Instead, this suggests that the leaders we will need are different in nature, that is, leaders who understand not only how humans function, but also how AI operates2.
As Van Quaquebeke and Gerpott put it, “They won’t be leading the humans within an organisation but leading the machines that lead the humans.”
Ethical stewardship – the future of leadership?
Human leaders will need to possess AI literacy to guide, prompt, and supervise AI-driven leadership systems effectively 5. A first step, then, is identifying ways to assess objectively whether human leaders understand the fundamentals of AI and its potential benefits and harms for employees6.
That is, leaders will increasingly focus on governing the systems that lead people, deciding the limits of optimisation, setting boundaries, and making ethical trade-offs.
Leadership is likely to be redefined as ethical stewardship.
As leadership becomes more automated, leadership systems will not only prioritise organisational goals such as system performance, but will also become more human-centred, with a stronger orientation toward employee interests, including well-being7. In such a landscape, empathy, moral imagination, and creativity become especially valuable, as these are qualities that machines cannot yet directly replicate. These are the very characteristics that will matter most as organisational procedures become increasingly complex7.
What skills will future leaders need?
Drawing from this, we can predict that human leaders will need to become especially skilled at recognising, explaining, and shaping the complex patterns that emerge when humans and AI work together. Additionally, as AI tools increasingly diffuse and redistribute decision-making processes, human leadership will shift toward balancing and orchestrating autonomy in relation to algorithms, feedback systems, and attentional demands.
At its core, however, the most important task for human leaders will be to assert and defend ethical judgement in relation to machine-driven systems7. As these systems become more deeply integrated, it also becomes more crucial than ever for human leaders to articulate and uphold ethical standards against the backdrop of increasingly powerful algorithms2.
Enabling algorithms as our moral agents could bring unintended consequences, especially if the data they are trained on reproduces existing biases. If human leaders understand even the basic functioning of AI systems, including how data or developers may misuse such systems to reproduce existing biases and values, they will be better positioned to meaningfully guide and shape their work environments2.
This gap between calculation and care is where human leadership still belongs.
The task ahead isn’t to defend old hierarchies or to fear automation, but to redefine leadership as a form of ethical design. It’s about ensuring that our systems, however intelligent, remain accountable to the people they serve.
More specifically, this will require human leaders to develop a kind of digital backbone, so to speak, that enables them to remain firm when technologies generate ethically questionable recommendations , such as disproportionately targeting certain groups for dismissal based on performance metrics or promoting constant AI integration without reflection on what this means for people and its broader implications8.
Rather than primarily motivating employees or instilling inspiration, human leaders will need to determine:
- which goals AI systems should optimise
- which organisational values must be preserved
- where and when automation should be curtailed, even if doing so reduces process efficiency.
This would mean being accountable for even those decision-making systems that are otherwise not so transparent, and overcoming the urge to hide behind the algorithm’s decision-making.
The leadership role will change
To revisit the question posed above, what is left for human leaders, we can expect a shift from telling people what to do toward directing the systems that shape how work is performed.
Human leaders will need to become especially skilled at recognising, explaining, and shaping the complex patterns that emerge when humans and AI work together.
As AI increasingly mediates decisions related to hiring, performance, development, and support, human leaders will become accountable for what these systems normalise. They will be responsible for ensuring that empathy is not reduced to a metric, fairness is not distorted by biased data, and diverse, often marginalised employee groups are not inadvertently excluded by the logics of optimisation5.
This future, too, has risks. It could privilege an elite of AI-literate experts and widen the gap between those who understand the systems and those who are governed by them.
It only works if it remains grounded in empathy and inclusion, and if it listens as much as it codes. Most critically, leaders must ensure that AI does not reproduce a single model of the ideal worker, one that is implicitly aligned with consistent availability, heteronormativity, and dominant career trajectories, at the expense of differences3.
Inclusive leadership, in this context, lies in the ongoing ethical work of keeping algorithmic systems receptive to diversity, complexity, and plurality in organisational life.
Leadership will therefore become less about the forms of presence or persuasion leaders once embodied, and more about the ethical boundaries they set, grounded in safeguarding inclusion and taking responsibility for how algorithms shape everyday work.
About the Authors
Vedika Lal is a postdoctoral researcher at the Institute for Leadership and Inclusive Management, attached to the Work & Human Relations department of ESCP Business School.
Zuzanna Staniszewska (PhD) is an assistant professor at Kozminski University in Warsaw and a research associate and visiting scholar in the Work and Human Relations Department at ESCP Business School in Paris.
Géraldine Galindo is full professor in the Work & Human Relations department on the Paris campus and director of the Institute for Leadership and Inclusive Management at ESCP Business School.
