Posted on LinkedIn April 20, 2026
Nearly half of organizational leaders now report piloting or fully implementing generative AI in their day-to-day workflows. (Russell Reynolds Associates, “Optimistic, with Exceptions: Leaders’ Views on Generative AI in 2025,”) The conversation in most boardrooms has shifted from “should we” to “how fast.” That shift is clarifying something most organizations weren’t ready to see.
When AI takes over synthesis, pattern detection, scenario modelling, and status reporting, the tasks that used to pass for leadership contribution in a lot of organizations, what remains is a visible gap between leaders who understand how organizations actually work and leaders who manage activity. AI didn’t create that gap. It made it harder to obscure.
What the research is telling us
In March 2025, MIT Sloan published research that moves beyond the standard “jobs at risk” framing. The study identified five uniquely human capability groups, Empathy, Presence, Opinion and Judgment, Creativity, and Hope, that AI cannot replicate, and found that human-intensive tasks involving these capabilities increased between 2016 and 2024. (MIT Sloan School of Management, “New MIT Sloan research suggests that AI is more likely to complement, not replace, human workers,” March 2025)
Presence is the one worth examining closely.
Not presence as a personality trait. Not charisma or executive gravitas or how you enter a room. Presence as a leadership capability: the ability to read what is actually happening in an organization, the political texture, the unspoken resistance, the commitment that got stated in the meeting and will quietly dissolve by Friday, and respond in a way that moves the situation forward.
Research from EHL puts it plainly: AI cannot sense when the same words will land differently with different people. It cannot read the political and cultural texture of a room and decide, in that moment, what the right thing to say is. (EHL Hospitality Business School, “WIL 2026 Insights: What AI Cannot Lead,” April 2026 ) That observation sounds straightforward. The operational implications are not.
Three places where presence drives outcomes, AI cannot touch
Signal interpretation before the data resolves.
AI produces analysis. It does not produce judgment about which signals matter right now, before the full picture is clear. The leader who can detect that a coalition is forming against a decision before anyone names it, or that a team’s compliance is not the same as their commitment, is operating on organizational intelligence that doesn’t surface in dashboards. That capability develops over years of paying close attention to how people behave under pressure. It is not a personality trait. It is a skill built through experience in complex organizations.
Decision quality in ambiguous conditions.
The highest-value decisions are not data-rich. They involve competing priorities, incomplete information, and a leadership team that is not fully aligned. The leader who can sit in that ambiguity, surface the real tradeoff without manufacturing false certainty, hold the tension long enough for the right decision to become visible, and move a group to genuine commitment rather than grudging compliance, is exercising a structural capability. It is also uncommon. Most leadership development models under-invest in it because it is harder to measure than throughput.
Execution that holds when conditions change.
Strategy rarely fails in the presentation. It fails in the translation from decision to implementation when context shifts, and the people carrying the work need a leader who can reorient them without losing the thread. That requires understanding the human dynamics of an organization well enough to know what is driving the drift and address it at the source. AI can flag that execution is off track. It cannot do what happens next.
The structural implication
Research in the Journal of Leaderology and Applied Leadership is direct on this point: tasks involving embodied presence, real-time relational trust calibration, and adaptive judgment are not replicable by AI. (Journal of Leaderology and Applied Leadership, “Artificial Intelligence and the Future of Leadership Development,” February 2026) The leadership capabilities that AI cannot replicate are precisely the ones that determine whether strategy gets implemented, whether teams stay aligned under pressure, and whether organizations can adapt when the environment changes.
Those capabilities don’t develop through a certification program. They develop through sustained attention to how organizations actually behave: the gap between what’s decided and what happens, between what’s stated and what’s true, between compliance and commitment. Leaders who have built that observational muscle over the years in complex organizations are not at a disadvantage in an AI-augmented environment. They are carrying a structural asset that becomes more valuable as everything below it gets automated.
The question is more specific than whether presence matters. Where in your leadership practice is presence already operating, and where are you still relying on activity to do the work that only presence can do?
Sources
Russell Reynolds Associates, “Optimistic, with Exceptions: Leaders’ Views on Generative AI in 2025.” Available at: russellreynolds.com/en/insights/articles/leaders-views-on-generative-ai-in-2025
MIT Sloan School of Management, “New MIT Sloan research suggests that AI is more likely to complement, not replace, human workers,” March 17, 2025. Available at: mitsloan.mit.edu/press/new-mit-sloan-research-suggests-ai-more-likely-to-complement-not-replace-human-workers
EHL Hospitality Business School, “WIL 2026 Insights: What AI Cannot Lead,” April 2026. Available at: research.ehl.edu/news-listing/human-centered-leadership
Journal of Leaderology and Applied Leadership, “Artificial Intelligence and the Future of Leadership Development,” February 2026. Available at: jala.nlainfo.org/artificial-intelligence-and-the-future-of-leadership-development
