The Next Leadership Assessment Will Include AI
Anna Drobakha, Global Digital Business & AI Transformation Director at Groupe SEB.
For decades, we have assessed leaders by what they know, how they think, how they lead people and what results they produce. We look at strategic judgment, functional expertise, emotional intelligence, communication and commercial performance. But the unit of leadership performance is beginning to change.
Executives increasingly work through AI assistants, intelligence engines and automated workflows. AI is becoming part of how leaders prepare decisions, read markets, challenge assumptions and manage their own attention. That raises a question boards and CHROs will soon have to confront: Should we keep assessing leaders only as individuals, or also assess the quality of the human–AI systems they build around themselves?
Fluency Is Becoming The Baseline
The first phase of enterprise AI was about access. The second has been about literacy: helping people understand what AI can do and where the risks lie.
That still matters. Demand for AI fluency in European job postings increased roughly fivefold between the fourth quarters of 2023 and 2025, according to the McKinsey Global Institute, and the EU AI Act now requires organizations deploying AI systems to ensure an appropriate level of AI literacy among the people using them.
But literacy is not capability. A leader can understand generative AI, know its limits and use it daily, yet still be unable to redesign a decision process or turn AI use into measurable value. We would never judge a leader’s financial capability only by asking whether they can read a balance sheet; we look at how they apply it to investment, risk and performance.
AI is no different. The question is shifting from “Does this leader understand AI?” to “Can this leader create value through AI—effectively, responsibly and repeatably?”
The Emerging Unit Of Performance
Leaders are starting to work through systems that monitor developments, synthesize information, prepare decision briefs, test assumptions and coordinate recurring workflows. The leader stays accountable, but the system through which they create value is now hybrid. The unit of performance is no longer simply the individual. It is the leader plus the AI, the context, the workflow and the governance around it.
That does not mean rewarding whoever has the most sophisticated technology. Access to premium models and enterprise data varies widely, and grading someone on the apparent quality of their tools would reward privilege, not leadership. The real question is how well a leader designs and manages the relationship between human judgment and machine capability.
What We Should Actually Assess
An AI-augmented leader is not the person who produces the fastest answers. The capability is broader, and most of it is judgment.
It starts with strategic framing: knowing where AI can create real value and where it should not be used at all, beginning from the business problem rather than the tool. It depends on context, because AI performance rises and falls on the goals, constraints and evidence a leader gives it, which makes providing context a form of system design, not just communication. It requires intelligent delegation between people and machines, judged by the same standard as delegation to people: clear purpose, defined boundaries and visible accountability.
As AI makes execution faster, judgment matters more, not less. The strongest leader may not be the one who accepts the most output, but the one who knows what to reject and who can catch a plausible, confident, incomplete answer before it becomes a decision.
That also requires responsible control. Leaders cannot simply hand questions of risk, ethics and accountability to legal or IT. Finally, there is value realization. Usage is not value. The measure is not how often AI is used, but whether the work gets meaningfully better as a result.
From Self-Reported To Demonstrated
Most leadership assessment blends interviews, self-evaluation, feedback and observed behavior. AI-augmented leadership will need the same multidimensional approach: practical, observed and evidence-based rather than a multiple-choice test of terminology.
Picture a simulation in which an executive must prepare a recommendation on an unfamiliar issue. What matters is not the final slide, but the process: how they framed the problem, what context they gave, what they delegated to AI, which sources they trusted, what they tested, what they rejected and which decisions stayed human. That reveals far more than any quiz.
In my own work with executives, the meaningful shift rarely happens when someone learns another set of prompts. It happens when they rebuild a real operating rhythm around AI: how they prepare for decisions, synthesize intelligence and challenge assumptions. That is the difference between learning about AI and learning AI by doing.
Getting It Wrong
Organizations should move carefully, because this can reward the wrong behavior. More automation is not more maturity: a leader who declines to automate a sensitive people decision may show better judgment than one who automates everything in reach.
Assessment also has to separate three things—the capability of the leader, the capability of the AI system and the conditions the organization provides—or leaders get penalized for weak infrastructure and rewarded for access they did not build. And its first home should be development and capability-building, not recruitment or promotion. Higher-stakes use will demand transparent criteria, standardized conditions, privacy safeguards and real human review.
Accountability Stays Human
Employers expect 39% of workers’ core skills to change by 2030, according to the World Economic Forum’s Future of Jobs Report 2025, with both AI capabilities and human leadership skills rising together. That combination is the point. The future is not about replacing human leadership with AI; it is about understanding how leadership changes when intelligence and execution can be augmented.
Boards may need to ask whether weak executive AI capability is now a strategic risk. CEOs may need to decide which roles require augmented capability first. And leaders may need to show not only what they personally know, but what they can responsibly achieve through the systems they build around themselves.
AI may prepare the analysis, coordinate the workflow and accelerate the execution. But the operating system is becoming hybrid while accountability stays human, and that is exactly what the next assessment will have to measure.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
