By Massimiliano Ferraris
Boards face structural instability driven by technological acceleration and geopolitical fragmentation. The primary governance vulnerability lies not in compositional diversity but in a deficit of integrative architecture. Resilient governance in non-linear environments requires redesigning decision structures to sustain coherent cross-domain integration, rather than relying on vertical specialization alone.
Rising uncertainty has not broadened the cognitive horizon of boards. In many cases, it has narrowed it.
The prevailing reaction has been defensive. Boards have strengthened their reliance on consolidated experience and increasingly favoured profiles that have already held apex roles, with linear and recognisable career paths aligned with traditional governance patterns.
Appointment dynamics often reveal a tendency toward endogenous reproduction of decision-making elites. Similar backgrounds, similar educational paths, similar professional socialisation, often the same managerial generation. In an environment perceived as unstable, experience is treated as a proxy for adaptive capacity, and familiarity becomes a mechanism to reduce uncertainty.
Cultural, cognitive, professional and international diversity frequently remain secondary. They are more visible in governance codes than in the actual mechanisms through which boards select members and operate. Even when new competences are formally introduced, they tend to be confined to specialist or advisory roles without materially influencing the overall decision architecture.
This produces a structural paradox. As the external environment becomes more heterogeneous, interdependent and non-linear, boards often become more cognitively homogeneous. Uniformity of experience and decision language reduces productive tension, compresses interpretative pluralism and reinforces forms of apparent consensus that slow the recognition of weak signals.
The limitation is not individual competence. It lies in the systemic structure through which competences interact. Decision quality depends less on who sits at the table and more on how much cognitive variety the governance system can absorb without fragmenting. Without that variety, the board risks functioning as a chamber that amplifies past patterns precisely when the future requires interpretative discontinuity.
One of the most persistent misunderstandings in contemporary governance is the belief that diversity is primarily a matter of composition.
In many organisations, diversity is treated as a static attribute, measurable in terms of gender, age, nationality or declared expertise, rather than as a dynamic property of the decision process. The outcome can be a formally heterogeneous board that remains substantively uniform at the decisional level, where individual differences do not translate into interpretative variety.
Faced with rising systemic complexity, boards have not significantly expanded their cognitive perimeter. Part of the reason is structural. The market offers very few genuinely multidisciplinary profiles. Individuals capable of integrating strategy, finance, law, technology, regulation and geopolitical analysis are rare. Dominant career trajectories remain vertical and specialised, oriented toward depth within a single domain rather than synthesis across domains.
As a result, boards may be composed of highly competent individuals who nonetheless share similar mental models, similar decision languages and similar approaches to risk.
Selection and co-optation mechanisms reinforce this pattern. Cross domain careers are often difficult to position within traditional skills matrices and may be perceived as lack of focus rather than integrative capacity. Homogeneous boards tend to select profiles that feel reassuring and recognisable, further reducing space for atypical figures. Even when diversity is invoked, it frequently remains peripheral, while the core decision nucleus continues to be shaped by similar trajectories and interpretative frameworks.
The central issue is not a deficit of talent. It is a structural deficit of integrative capacity at the apex of decision making. In non-linear environments, the weakness of boards does not lie in the quality of individual members but in the scarcity of figures able to operate as connectors across complex domains, translating heterogeneous signals into coherent choices. The main fragility of contemporary governance sits in this gap between specialisation and integration.
The emergence of artificial intelligence, and particularly autonomous agents, makes this gap more visible and more consequential.
AI accelerates decision cycles, reduces the cost of processing information and multiplies the number of available signals. At the same time, it increases interdependence among domains that were previously treated separately. Technology, strategy, finance, law, compliance, security and reputation converge within the same decision.
In this context, vertical specialisation, historically a strength of boards, can become a source of fragility. AI systems do not generate isolated technical questions. They generate hybrid trade-offs. A model may be efficient but legally non-compliant. High performing but exposed to reputational risk. Acceptable in one jurisdiction and problematic in another. Advantageous in the short term yet destabilising in the medium term.
Breaking these trade-offs into functional silos does not simplify the decision. It distorts it.
Boards are increasingly required to govern technologies that act across the organisation, yet they often lack sufficient figures capable of integrating technological understanding, regulatory implications, financial impact and strategic consequences within a single frame. Delegating this integration to committees or external advisors weakens systemic accountability and introduces latency at the very moment when AI reduces the time available to decide.
When AI agents move from supporting decisions to executing them, interacting directly with clients, suppliers, markets or digital infrastructures, the boundary between oversight and action becomes thinner. Governance can no longer rely only on ex post controls or static policies. It must anticipate scenarios, define dynamic limits and understand the systemic implications of technological choices.
AI therefore highlights a widening gap between the complexity of decisions boards are asked to make and the cognitive structure of the profiles that compose them. The risk is not simply technological error. It is progressive loss of effective control as systems gain autonomy.
The growing reliance on consolidated experience should not automatically be read as prudence. Under conditions of radical uncertainty, past experience is not always a reliable predictor of future capacity, particularly when contextual parameters change faster than institutional learning cycles.
When experience becomes the dominant selection criterion, the available cognitive variety tends to shrink. Experience that is not continuously refreshed can crystallise interpretative schemes that were effective in prior contexts but become less informative in environments shaped by technological acceleration, global interdependence and asymmetric shocks. In such circumstances, experience may function less as adaptive capital and more as inertia.
The relevant question is therefore not whether boards have enough experience, but whether they have experience capable of learning, unlearning and recombining knowledge across domains.
This is the function of what can be described as a bridge role (n what follows, I use the term “bridge manager” to describe a stable integrative role within the architecture of the board — not an additional specialist, but a function capable of translating across strategic, legal, technological, and financial domains. The bridge manager does not add vertical expertise. It enables integration across domains that would otherwise remain cognitively fragmented. Without such integration, boards slow down not because they ignore change but because they interpret it through partial lenses that fail to capture its systemic nature).
Not an additional specialist, but a connector. Someone able to render otherwise isolated competences communicative, transforming vertical expertise into integrated decision capacity. Without such integration, boards slow down not because they ignore change but because they interpret it through partial lenses that fail to capture its systemic nature.
Many boards continue to operate as confederations of specialists. Each member may be excellent within a given domain, yet the absence of a shared integrative frame makes it difficult to address issues that cut across multiple dimensions simultaneously. In linear environments, this configuration can be efficient. In complex and interdependent contexts, it produces friction. Information accumulates without translating into decision. Trade-offs remain implicit. Responsibility fragments. Decisions are postponed or diluted.
What is required is not simply more coordination, but a higher level of integration in the architecture of decision making.
By 2026, national and transnational dimensions are no longer separable layers of analysis. Technology, capital, data, markets and regulation move according to different logics yet generate simultaneous effects. Geopolitical trajectories are not background conditions. They are embedded in corporate choices.
Boards often respond by decomposing complexity into manageable domains. AI is treated as a technological issue. Digital platforms as communication tools. Markets as implicitly domestic. This fragmentation makes individual elements more tractable but weakens the ability to grasp their systemic interaction.
The problem is not lack of information. It is lack of connection.
In this environment, international and cultural diversity acquire a different meaning. They are not primarily reputational signals or compliance exercises. They are elements of decision security. Boards composed exclusively of domestically formed profiles may interpret risk symmetrically and underestimate asymmetric shocks that unfold across jurisdictions and value chains.
International perspectives introduce interpretative discontinuities that can strengthen resilience. They expand the range of cognitive frames within which decisions are situated. When such plurality is accompanied by effective integration, diversity becomes a multiplier of strategic stability rather than an accessory attribute.
Continuity remains an important governance value. It preserves institutional memory and coherence. However, continuity without cognitive renewal can turn into lock in. Even in the presence of formal turnover, new members may be absorbed into existing interpretative patterns rather than reshaping them. Governance may continue to function procedurally while progressively losing alignment with an external environment that evolves faster than its internal categories.
Another underestimated fragility lies in excessive internal bureaucratisation. Extended approval chains, proliferating committees and procedures designed primarily to dilute responsibility can generate chronic slowness. In stable environments, this complexity may appear neutral. In unstable environments, it becomes a strategic risk.
The challenge for boards is not to choose between speed and control. It is to redesign decision flows so that traceability, accountability and regulatory compliance are preserved while unnecessary latency is reduced. Contemporary technologies allow this. Digital workflows, automated audit trails and advanced tracking systems make it possible to simplify processes without weakening oversight.
Redesigning decision processes, however, is not merely an operational improvement. It is an architectural choice about how power and responsibility are structured. It requires deliberate intent at the top.
The central issue is no longer whether boards should update their skills or open themselves to new competences. It is whether they are willing to operate with decision architectures designed for stability in a context that is structurally unstable.
A board that does not incorporate a stable integrative function implicitly accepts several risks. Slower critical decisions. Defensive bureaucratisation. Loss of cognitive continuity during transitions. These are not abstract possibilities. They are observable patterns.
The decisive governance question therefore concerns not only who sits on the board, but how the board is structured to decide over time. Introducing a connective function capable of traversing competences, jurisdictions and decision cycles is not optional enrichment. It is institutional risk mitigation.
The final question is not whether a board possesses experience, diversity or excellent competences. It is whether it has consciously chosen an architecture capable of absorbing systemic instability and transforming it from a source of disorder into a lever of strategic governance.
In structurally unstable environments, governance cannot avoid architectural choice. Even inaction reshapes the architecture.
Massimiliano Ferraris
