Executive succession when roles are changing faster than job descriptions

Why static role profiles are losing their value

Stanford’s 2026 AI Index shows that AI capability and adoption continue to expand quickly, while organisational use is already widespread ¹. For succession, the practical implication is straightforward: the next incumbent may inherit a mandate that changes midstream, especially where AI alters how work is structured or how decisions are made. A profile built around yesterday’s accountabilities can be too narrow to identify leaders who can operate in ambiguity.

This does not mean abandoning role design altogether. It means treating the position description as a working hypothesis rather than a fixed map. Boards should distinguish between what must remain stable, such as stewardship, risk ownership and stakeholder trust, and what is likely to evolve, such as operating rhythms, decision rights and the mix of human and digital capability needed around the role.

What to look for instead of a perfect fit

Harvard Business School’s AI Institute reports that in a controlled study, leaders who performed well with AI teammates also tended to perform well with human teams ². The study is a laboratory exercise, not a direct blueprint for board succession, but it offers a useful signal: leadership capability can travel across contexts when the underlying task is information gathering, coordination and judgment. In practical terms, succession panels should pay attention to how candidates ask questions, draw out views and update their thinking when the evidence changes.

An executive who can succeed only when the environment is familiar may be less useful than one who can lead through incomplete information. Look for people who are comfortable testing assumptions, who can simplify without oversimplifying, and who can build alignment without forcing false certainty. These are not new virtues, but they become more important when the role itself is moving.

How AI changes the succession conversation

MIT Sloan distinguishes between broad, organisation-wide generative AI tools and more tailored solutions built for particular business needs ³. That distinction matters for succession because the capabilities required in the next executive may vary depending on whether AI is being used mainly for personal productivity, for process redesign, or for strategic transformation. One-size-fits-all “digital fluency” is therefore too vague to guide selection.

MIT also notes that leadership for AI work needs strategic direction, collaboration, experimentation and integration across business units ⁴. For succession planning, that suggests a useful frame: identify whether the role will require a translator, an operator, a sponsor or an architect of change. Sometimes it will need all four, but not always in the same person or to the same degree. The more explicit the board is about the likely shape of change, the better it can assess whether a candidate has the right range, not just the right résumé.

Practical steps for boards and CEOs

First, build succession scenarios around different futures rather than one forecast. Ask what the role looks like if AI is adopted conservatively, selectively or across the enterprise. Then test each successor candidate against those versions of the job. This is more demanding than matching a person to a single incumbent’s profile, but it is closer to reality.

Second, use case-based assessments that require candidates to prioritise, delegate and explain trade-offs under uncertainty. Harvard’s study points to the value of questioning and turn-taking in leadership ². In a succession context, that means observing how candidates surface hidden information and how they respond when advice conflicts.

Third, broaden the evidence base. Include stakeholders who understand operations, technology, culture and risk, but keep the decision disciplined. Too many voices can produce consensus around familiarity rather than readiness for change.

Fourth, treat transition support as part of succession, not an afterthought. A leader stepping into a fluid mandate will need early clarity on where they can shape the role and where they must preserve continuity.

A more resilient succession discipline

The most useful succession plans will be those that accept uncertainty rather than pretending it can be removed. Research from these university sources suggests that leadership capability, AI governance and organisational change are tightly connected ¹³⁴. My advisory perspective is simple: when a role is evolving faster than its title, boards should select for learning speed, judgment and the ability to organise people around a changing agenda.

That is a more demanding standard than checking whether someone has done the job before. But it is also a fairer one, because it asks whether the person can help define what the job must become.