Navigating foundational uncertainty: a dynamic-adaptive model of enterprise digital transformation in the age of generative AI

Why this matters for everyday work

Many people sense that generative AI is changing work faster than companies can plan. This paper asks a simple but urgent question: what happens to business strategy when the most important technology is powerful, unpredictable, and poorly understood even by its creators? Instead of treating AI as just another tool to roll out with a roadmap, the author argues that firms must reinvent how they learn, decide, and compete in a world where the ground keeps moving beneath their feet.

From tidy plans to shifting ground

For decades, management thinking assumed that new technologies unfolded in a mostly predictable way. Companies watched from the sidelines, waited until a tool was mature and well understood, and then ran a planned “digital transformation” project to put it to work. In this older view, firms combined their resources, chose suitable technologies, and executed carefully designed programs to gain a stable edge over rivals. Technology itself was seen as a knowable, controllable ingredient in an otherwise rational recipe for success.

When the core technology stops behaving

Generative AI breaks this comfortable story. Its abilities can appear suddenly as models grow larger, and its inner workings are so opaque that even experts struggle to explain how it reaches its answers. Companies adopt AI systems whose future skills, weaknesses, and side effects cannot be forecast in a simple straight line. This creates what the paper calls “foundational uncertainty”: the problem is not just missing data, but that the usual methods for predicting and planning no longer work. Instead of a stable ladder of progress, managers face an evolving, partly mysterious partner that can surprise them in both good and bad ways.

Learning by doing, not by predicting

In this new reality, the paper argues, companies must flip the usual order of thinking first and acting later. Because no one can fully grasp in advance what generative AI will do inside a specific business, firms need to probe, experiment, and then make sense of what they discover. The author calls this “adaptive sensemaking”: small, repeated trials where people watch how AI behaves in real work, uncover unexpected uses and failure modes, and gradually build a working understanding. Strategy stops being a fixed plan on paper and becomes an evolutionary process that grows out of these lived experiences with the technology.

The push and pull inside the organization

This constant experimentation creates tension within the firm. Senior leaders want order, control, and reliable results; frontline teams working with AI uncover messy, surprising possibilities that do not fit existing rules. The paper suggests that this friction is not a problem to eliminate but the very engine of progress. Top-down pressure pushes experiments toward business goals, while bottom-up discoveries force the organization to adjust its structures, processes, and even goals. Managed well, this cycle of “contention and adaptation” lets companies both exploit what they already know and explore what AI makes newly possible.

Turning adaptability into the real advantage

Because powerful AI tools and data are increasingly available to many players, no single model or dataset can guarantee long-term superiority. Instead, the paper claims that the true winning asset is a “dynamic adaptive capability”: the organization’s ability to keep reconfiguring how it learns, experiments, and changes course. In plain terms, the lasting advantage is not owning the best tool, but being the best at learning from that tool and reshaping the business in response. Companies that build this muscle of continuous adaptation can string together many short-lived advantages, staying ahead in an environment where every specific edge quickly erodes.

What this means going forward

For a lay reader, the main takeaway is that generative AI forces companies to trade rigid long-term plans for flexible learning loops. The paper concludes that firms should treat AI less like a finished product to install and more like a powerful, unpredictable collaborator that must be explored with care. Those organizations that invest in safe experimentation, open discussion of failures, and a culture that welcomes change are most likely to thrive. In the age of generative AI, surviving and succeeding is less about mastering the machine once, and more about building the capacity to keep learning with it, over and over again.

Citation: Lu, T. Navigating foundational uncertainty: a dynamic-adaptive model of enterprise digital transformation in the age of generative AI. Humanit Soc Sci Commun 13, 451 (2026). https://doi.org/10.1057/s41599-026-06689-z

Keywords: generative AI, digital transformation, organizational learning, business strategy, competitive advantage