When AI Stops Being a Tool and Starts Thinking With Us: The Hidden Shift Reshaping SME Strategy

A quiet but fundamental shift is underway in how small and medium-sized enterprises (SMEs) formulate strategy. For decades, strategic management assumed a clear division of labour: humans interpreted markets, humans made decisions, and technology merely supported analysis. The emergence of artificial intelligence (AI), however, is gradually dissolving that boundary.

New research published in Administrative Sciences by Grant Freedman introduces a novel conceptual framework called Strategic Edge Architecture (SEA), which explains how AI reshapes strategic decision-making in SMEs. For SMEs, strategic decision-making is often concentrated in a small group of owner-managers working under conditions of uncertainty, limited analytical resources, and constant environmental pressure. The SEA framework suggests that these constraints are becoming more consequential, not because the environment is changing, but because the volume, velocity, and complexity of signals demanding attention now exceed unaided human cognitive capacity.

AI systems alter this equation by expanding what firms can detect and process. They continuously scan markets, identify weak signals, and surface patterns across fragmented data sources that human decision-makers would struggle to integrate in real time. Still, the point is not simply faster analysis. It is the restructuring of attention itself: what organizations notice, what they ignore, and what they consider strategically relevant is increasingly shaped by machine-generated interpretation layered into managerial cognition.

From Data Overload to Strategic Foresight

AI strengthens environmental sensing by widening the range and granularity of signals firms can capture. This enhanced sensing feeds into what the framework calls "prospective sensemaking," a forward-looking form of interpretation in which leaders do not merely explain what has happened but actively construct plausible futures. In traditional strategy theory, sensemaking has often been retrospective, anchored in interpretation of past events. SEA extends this logic by arguing that AI enables organizations to simulate, compare, and evaluate multiple future trajectories with far greater depth and speed than human cognition alone allows. Strategic decision-making thus shifts from reacting to disruptions toward anticipating them, particularly in areas such as regulatory change, competitive dynamics, and sustainability transitions.

The research insists that AI does not create strategic advantage by itself. Instead, its value is conditional on how it is governed and integrated into human decision authority. The concept of human-in-the-loop (HITL) governance becomes critical here, functioning as the mechanism that preserves interpretive control, contextual judgment, and accountability within AI-augmented systems.

Without such governance, organizations risk a subtle but powerful form of dependency in which decision-makers begin to defer excessively to algorithmic outputs. This phenomenon, often described as automation bias, can narrow rather than expand strategic thinking, particularly in SMEs where technical expertise may be unevenly distributed and leadership structures are highly centralized. In this view, AI does not simply enhance cognition, but can also reshape authority, potentially displacing human judgment if not carefully managed.

The New Competitive Edge Is Cognitive Infrastructure

Competitive advantage is moving away from traditional resource-based explanations and toward something more structural: the design of organizational thinking itself. SEA introduces the idea of "cognitive infrastructure" to capture this shift - the integrated system of human cognition, AI capabilities, and governance mechanisms through which interpretation and strategy are produced. This infrastructure determines not only how efficiently firms process information but how they construct meaning under uncertainty. As a result, two firms with similar access to AI tools may diverge significantly in performance depending on how deeply those tools are embedded into decision processes and how effectively human oversight is maintained. In other words, advantage no longer comes from having AI, but from how coherently AI is woven into organizational cognition.

For SMEs in emerging economies, AI adoption is accelerating but often occurs in fragmented or informal ways. The research suggests that simply introducing AI tools is unlikely to generate meaningful strategic gains unless firms simultaneously develop governance routines, interpretive discipline, and the ability to critically evaluate machine-generated insights.

Additionally, the framework highlights a tension: while AI can significantly enhance strategic foresight and sustainability integration, helping firms identify environmental, regulatory, and ESG-related risks and opportunities, it can also amplify vulnerabilities if deployed on poor-quality data or within weak institutional contexts. Risks such as algorithmic bias, data poverty, and overreliance on historical patterns can distort strategic perception and lead firms toward suboptimal or overly conservative decisions.

Sustainability at the Heart of Strategic Thinking

The research integrates sustainability into the core of strategic cognition rather than treating it as an external compliance requirement. In the SEA framework, environmental and social signals are not separate from strategy; they are embedded within the same cognitive system that processes competitive and technological information. Sustainability is not something firms "add on" to strategy later, but something that emerges through the same interpretive processes that shape competitive positioning.

AI systems, when properly designed and governed, can therefore help SMEs align business decisions with sustainability transitions in real time, identifying circular economy opportunities, regulatory shifts, and environmental risks that would otherwise remain outside managerial attention.

The study challenges long-standing assumptions in strategic management about where cognition resides. Traditional theories locate decision-making in individuals or organizational routines. SEA instead distributes cognition across a network of humans and machines, suggesting that strategic capability emerges from interaction rather than location. The redefinition implies that improving strategic performance is not only a matter of hiring better managers or acquiring better data, but of redesigning the architecture through which interpretation itself occurs.

Shared Intelligence and the Future of Strategy

The framework remains conceptual and requires empirical validation across different SME contexts and institutional environments. It also assumes a baseline level of digital infrastructure and data availability that may not be present in all emerging economy settings. Moreover, some of its constructs, such as prospective sensemaking and cognitive infrastructure, are theoretically rich but operationally challenging, raising questions about how they can be reliably measured in practice.

The framework also acknowledges a broader cultural limitation: much of its theoretical foundation draws on Western organizational and strategic traditions, which may not fully capture alternative decision-making logics shaped by relational or indigenous knowledge systems.

Strategy is becoming less about isolated human judgment and more about how effectively organizations design systems in which human and machine intelligence interact. For SMEs, this shift is not optional, but structural.