The Uncomfortable Truth at the Top
According to a 2025 IBM and Censuswide survey of 3,500 senior leaders across Europe and the Middle East, sixty percent of organisations report a significant AI literacy gap at leadership level. The gap is not between executives who use AI and those who do not. It is between those who can genuinely evaluate, govern, and direct AI-driven decisions — and those who are performing fluency they do not actually possess. The distinction matters enormously. And it is now being investigated with academic rigour by researchers who are sitting, as students and practitioners, at exactly this intersection.
Research from the Inside: What Executives Actually Say About AI
Salvatore Scuderi has over twenty years of experience as a Senior Partner in Insurance at a global IT services and digital transformation company, he has led complex migration and transformation projects on a global scale, managed AI deployments across insurance, reinsurance, and healthcare insurance sectors, and served on boards that have had to make precisely the kind of AI-informed strategic decisions that most executive education programmes still do not prepare leaders for.
At the European Institute of Management, Scuderi is pursuing a PhD in Management with a research focus on AI-based hybrid intelligence in strategic decision-making — investigating, through direct engagement with senior executives, what AI literacy actually means in practice rather than in theory.
His findings are instructive — and in places, deeply uncomfortable. The research reveals that many executives initially associate AI competence with technical specialisation. This perception creates what Scuderi's interviews describe as psychological barriers — a reflexive stepping back from AI because leaders believe they are "not technical enough." As one senior executive in his study stated plainly:
"The moment people hear AI, they think they need to become data scientists — and that immediately creates resistance."
— Executive interview participant, Scuderi (2024/25)
But the barrier, the research shows, is not technical. It is conceptual. And closing it does not require executives to learn to build algorithms. It requires them to learn to govern them.
What AI Literacy Actually Demands of Leaders — According to Those Who Have It
Scuderi's research draws a sharp distinction between two failure modes at executive level: over-dependence and reflexive resistance. Both, the study finds, stem from the same root: a lack of genuine strategic clarity about what AI can and cannot do. The executives in his study who had developed genuine AI literacy — those who were integrating AI into strategic decision-making effectively and responsibly — described the competency not in technical terms but in judgement terms.
"I do not need to build the algorithm. I need to understand when I can trust it and when I should challenge it."
Another offered a formulation that cuts to the heart of the leadership question:
"AI does not replace executive judgement. It changes the quality and speed of the information leaders work with."
This is not a comfortable reframing for a generation of senior leaders who built careers on being the person in the room with the best information. The AI era does not reward that. It rewards the person who can most critically evaluate information they did not generate, from systems they did not build, producing outputs they cannot fully audit. That is a different skill. It requires development. And it is not what most MBA programmes are currently building.
Four Competencies That Define Genuine AI Literacy at Executive Level
Drawing on both Scuderi's research and the broader literature, four competencies emerge as the defining markers of genuine AI literacy at leadership level. They are worth examining carefully — not least because they expose the gap between what MBA programmes claim to offer and what many actually deliver.
1. Strategic judgement about AI deployment. This means understanding where AI creates genuine organisational value versus where it creates the appearance of efficiency — automating processes that were not worth doing in the first place, faster. Scuderi's participants described a shift from "technology enthusiasm" to a more grounded understanding, captured by one executive as:
"AI is not the strategy. It is the amplifier of the strategy."
— Executive interview participant, Scuderi (2024/25)
Leaders who cannot make this distinction are not equipped to evaluate AI investments, challenge vendor proposals, or align AI adoption with long-term strategic objectives rather than short-term cost optics. As another participant in the study noted:
"The real challenge is not implementing AI. The real challenge is changing how leaders think when AI becomes part of the decision process."
2. Critical evaluation of AI outputs. The ability to interrogate what an AI system produces — identifying hallucinations, recognising bias in training data, questioning the assumptions embedded in an algorithmic recommendation. MIT Sloan professors describe this as "leadership-level AI literacy": not technical mastery, but informed scepticism. Scuderi's study found that leaders who lacked confidence in their own AI understanding tended toward one of two dysfunctional extremes: excessive dependence on AI outputs, or blanket resistance. As one participant observed:
"The danger is not that AI gives wrong answers. The danger is that managers stop asking difficult questions."
Another captured the organisational consequence of this failure with precision:
"AI can identify patterns, but it cannot understand the political or emotional consequences of a strategic decision."
3. Governance and ethical accountability. The EU AI Act, now in full force, imposes fines of up to seven percent of global annual revenue for violations involving high-risk AI systems. Governance is no longer a compliance function — it is a board-level responsibility. Scuderi's research found this particularly acute: executives described fears around "accountability ambiguity" — uncertainty about who bears responsibility when an AI-supported decision fails. Boards are discussing AI investments, but many are still not discussing AI accountability. That gap is not a technology problem. It is a leadership problem.
4. Human-AI collaboration design. Understanding how to redesign workflows, decision processes, and reporting structures to integrate AI effectively — while preserving the human judgement that makes those processes accountable and adaptable. The dominant finding from Scuderi's interviews is captured in a single participant's formulation:
"AI gives speed and patterns. Humans give meaning, context, and responsibility."
— Executive interview participant, Scuderi (2024/25)
The Hybrid Intelligence Insight: Why AI Performance Is a Human Question
One of the most significant contributions of Scuderi's doctoral research is its insistence that AI performance in organisational contexts is not primarily a technology question — it is a human and decision-making question. This aligns with a notable signal from the World Economic Forum's 2026 meeting in Davos, which confirmed what Scuderi's research had already identified empirically: organisations create value when AI is designed to augment human judgement, not replace it. The AI dividend — the actual return on investment from AI deployment — emerges at the intersection of human engineering, trust, and decision context.
As Scuderi himself articulated ahead of the Data Center Hybrid Basel 2025 conference, where he spoke on Strategic Leadership in the Age of AI and Quantum Computing:
"It's about building a bridge between technology and humanity, guided not by control but by conscious responsibility. Because the future isn't made by machines — it's shaped by the leaders who dare to cross the bridge."
— Salvatore Scuderi, EIM PhD Candidate, 2025
This is what Scuderi means by hybrid intelligence: not a technical architecture, but a leadership philosophy. The hybrid intelligence model recognises that human intuition, contextual understanding, emotional intelligence, and ethical reasoning are not deficiencies to be automated away. They are the capabilities that give AI-driven decisions their accountability, their adaptability, and their legitimacy. The leader who internalises this model is not threatened by AI. They are empowered by it — because they understand their own irreplaceable role within it.
Why the MBA Cannot Afford to Get This Wrong
There is a structural problem that Scuderi's research exposes implicitly and the broader literature addresses directly: the AI literacy gap at executive level is, in significant part, a programme design failure in business education. MBA graduates who cannot engage critically with AI systems, who cannot evaluate an AI investment case, who cannot lead a governance conversation about algorithmic accountability — these are not unlucky individuals. They are the products of programmes that were not designed for the world their students would actually work in.
Adding an AI in Business module to an unchanged curriculum is not a response to this challenge. It is a performance of response. The AI literacy that today's leaders actually need is not built through a module. It is built through the kind of rigorous, interrogative, intellectually demanding education that develops genuine analytical depth across every dimension of the programme.
The design principles that make this possible are not complicated: tutorial-based learning that requires students to construct and defend original arguments under intellectual pressure; cross-disciplinary curriculum that develops the systems thinking AI requires leaders to have; assessment models that reward the quality of reasoning rather than the recitation of frameworks; faculty engagement that bridges academic rigour with applied professional reality.
Scuderi's own doctoral research is itself an example of this principle in practice. His investigation is not theoretical. It draws on direct engagement with senior executives navigating real AI decisions in real organisations. The knowledge he is generating is not a synthesis of existing literature — it is original, empirically grounded, and professionally actionable. This is doctoral education doing exactly what it should do in an era where AI can synthesise what is already known, but cannot determine what needs to be discovered.
What EIM's Approach Actually Builds
The European Institute of Management's MBA programme is built around the pedagogical principles that make genuine AI literacy possible — not as a stated ambition, but as a design consequence.
The Oxbridge-style tutorial model at the heart of EIM's approach requires students to develop, articulate, and defend original thinking in real time. There is no template answer. The tutorial develops precisely the critical reasoning capacity that Scuderi's research identifies as the defining characteristic of AI-literate leaders: the ability to interrogate, not just consume.
EIM's nine MBA specialisations — spanning financial services, healthcare management, logistics and mobility, manufacturing, renewable energy, telecommunications, agribusiness, tourism and hospitality, and professional services — ensure that AI literacy is developed in specific industry contexts. An MBA graduate specialising in financial services who understands AI governance in the context of algorithmic trading, fraud detection, and credit scoring models is not carrying generic AI knowledge. They are carrying contextual expertise that is immediately deployable — and immediately defensible in a board conversation.
For professionals seeking doctoral-level engagement, the EIM PhD in Management and the Doctor of Business Administration represent the most rigorous pathways available. The DBA, in particular, is designed for the experienced professional who has encountered the AI challenge firsthand and is ready to generate original knowledge from that encounter — with the authority of deep professional expertise and the rigour of doctoral scholarship.
The Leadership Imperative
The AI literacy gap is real. It is growing. And it is not closing itself.
The executives who will lead effectively in 2030 are not those who waited until they fully understood the technology. They are those who developed, through rigorous education and disciplined practice, the strategic judgement, critical evaluation capacity, governance understanding, and collaborative intelligence that make AI a leadership instrument rather than a leadership threat.
As Salvatore Scuderi concluded from his research with senior leaders across industries:
"Executives do not need to become AI engineers. They need to become responsible orchestrators of human and machine intelligence."
— Salvatore Scuderi
The European Institute of Management's MBA and DBA programmes are designed to build exactly those leaders.