Author: Alice Jing SHAN, PhD, founder of ScholarLand Ltd. and an independent researcher working on complex systems through the lens of relational dynamics. Her work explores how relationships shape the stability and collapse of systems across scales over time, including how coherence is maintained under constraint and how risk and failure emerge. She combines long-horizon research with applied structural diagnosis and writes across disciplines.
Risk is increasingly understood not as an external condition, but as something that emerges within systems. This shift moves attention away from treating risk as an object “out there” to be measured, and toward recognising it as something that takes form through the dynamics of complex systems. Yet even within this shift, risk often remains at the level of description, without a clear account of the processes through which it takes form. We point to interdependencies, sensemaking, and boundary choices, but the question of how risk comes into being within a system remains less clearly articulated.
Relational coherence as a generative condition
One way of approaching this is to consider risk not as a condition, but as an emergent effect of changes in relational coherence. Systems do not simply contain risk. They generate it as the relationships that hold them together begin to shift under constraint, not only through their structure, but through how those structures evolve over time. As coherence weakens, a system’s capacity to coordinate, interpret, and respond may begin to fragment. What is later described as vulnerability or failure can often be understood as downstream expressions of these earlier relational shifts. From this perspective, risk can be understood as patterns of failure emerging along specific relational pathways.
How this becomes visible in socio-technical systems
This is particularly visible in contemporary socio-technical systems.
For instance, in AI-driven welfare or public service systems, risk does not arise from a single failure point. It often emerges gradually as relationships between data, decision-making processes, institutional constraints, and lived realities begin to drift out of alignment.
As systems scale, constraint such as efficiency, standardisation, and resource limitations reshape these relationships. Over time, the system may continue to function operationally, while its relational coherence weakens. What later appears as “bias,” “misclassification,” or “loss of trust” may already be downstream expressions of earlier shifts in how the system holds together. Different systems, under different constraints, will generate different trajectories of instability, particularly as the speed and scale of capability increase.
This suggests that managing risk may be less about identifying isolated threats, and more about understanding how coherence is maintained, or lost, within appropriate constraint over time, including how responses to risk can in turn reshape the conditions from which further risks emerge.
Implications for governance and engagement
In practice, this has implications for how risk is engaged with in governance contexts.
Where risk is approached primarily through categorisation, measurement, and mitigation, the focus tends to remain on identifiable events or outcomes. While these approaches are necessary, they may not fully capture how risk is being continuously generated within the system itself. A relational perspective draws attention to the conditions under which coordination becomes strained, interpretation diverges, and response capacity begins to fragment. These shifts are not always immediately visible, yet they shape how systems behave under pressure and how risks accumulate over time.
Engaging with risk in this way does not replace existing approaches, but situates them within a broader understanding of how system behaviour emerges under constraint.
Closing Remark: Risk as an emergent effect of relational conditions
This remains an area of ongoing exploration.
If risk is approached primarily as something to be identified and managed, rather than as something generated within systems, there is a risk of repeatedly addressing its manifestations while leaving its underlying conditions unchanged. Attending to how relational coherence forms, shifts, and weakens under constraint may therefore offer a complementary perspective for engaging with risk in complex systems, particularly where systems continue to function while their capacity to respond is already being compromised.