The Myth of Adoption
Thinking on the Transformation of the Enterprise: a note for strategy and transformation leaders
There is a familiar word making its way through nearly every conversation about artificial intelligence in the enterprise.
Adoption.
It is an ordinary word; one we have used for decades to describe how organizations introduce new technologies. We adopt software, systems, processes, and new ways of working. There are adoption plans, adoption curves, adoption metrics, and executives responsible for ensuring that adoption occurs.
The word is reassuring. It suggests that we know what kind of undertaking we are engaged in.
A technology arrives. We evaluate its usefulness, determine where it belongs, prepare people to use it, and measure the results. Eventually, the technology becomes part of everyday operations.
There is nothing troubling about this description. Much of our experience with enterprise technology fits it. But the language of adoption does more than describe what organizations are doing.
It establishes the undertaking itself.
Artificial intelligence enters organizational life at a level unusually close to practices we have long associated with thinking itself.
Once we name the arrival of an AI adoption initiative, a familiar world of action becomes available. We establish deployment plans, assign responsibility, train users, measure utilization, and organize our efforts around achieving adoption.
The language gives us a way to proceed. It also shapes what we attend to and what passes unnoticed. Consider the relationship implied by the word.
The organization is the active party. The technology is something it acquires and introduces into an existing environment. Leadership decides how it will be used. Established practices become faster or more capable. The organization remains, in some essential sense, the organization it was before. It simply possesses new capabilities.
Of course, technologies have always changed the organizations that employ them. The spreadsheet altered financial work. Databases changed how information was organized. Digital communication reshaped relationships across geography and time.
Still, artificial intelligence enters organizational life at a level unusually close to practices we have long associated with thinking itself. Writing. Interpretation. Analysis. Synthesis. Conversation. Judgment.
These are among the practices through which an organization understands its circumstances, forms commitments, exercises authority, and decides what to do.
The introduction of AI into these practices changes more than the means by which work is accomplished. It changes the conditions under which the work takes shape.
Consider the preparation of a strategic memorandum.
A manager is asked to examine a difficult business question and recommend a course of action. The work takes several days. Information is gathered, competing interpretations considered, assumptions challenged. An initial argument is written, reconsidered, and revised.
A conversation with a colleague changes the direction of the inquiry. A sentence resists being written because the author has not yet determined what it ought to say. What initially appeared to be the problem turns out to be a symptom of something else. Eventually, a memorandum emerges.
An AI system can now produce a polished version of that memorandum in minutes, and we are impressed. What required days can apparently be accomplished in moments.
The memorandum enters a leadership meeting.
When a convincing answer is available immediately, remaining with an unresolved question becomes harder to justify.
It contains a persuasive account of the problem, a coherent analysis of the alternatives, and a reasonable recommendation. The participants can examine it, challenge it, and decide what to do. Nothing prevents good judgment. But the interpretation is available before the conversation that would have produced it.
The meeting begins with a plausible answer before the participants have developed a shared understanding of the question. And because the answer is plausible, the difference can pass unnoticed.
Something has changed in the relationship between producing an answer and arriving at an understanding. That relationship deserves attention, particularly in organizations accustomed to treating friction as something to eliminate.
We simplify processes, remove unnecessary approvals, automate repetitive work, and make information easier to retrieve. Much of this is good management. Yet not everything that slows the work is an obstacle to its accomplishment.
Sometimes hesitation reveals that the question is poorly formed. Disagreement exposes an assumption that everyone had accepted. The difficulty of writing an argument becomes the occasion through which the author discovers what they actually think.
What looked like friction was sometimes formation.
When a convincing answer is available immediately, remaining with an unresolved question becomes harder to justify. When competent work can be produced in minutes, the time once devoted to developing an understanding begins to appear excessive.
No one needs to declare that reflection is unnecessary. The surrounding conditions simply make it more difficult to defend. As expectations of competent performance change, so do the practices through which people develop judgment.
The difficulty is not that adoption is the wrong word. It remains useful for describing the practical work of introducing AI into an enterprise. The difficulty arises when that description becomes our understanding of the event itself.
We imagine an organization standing on one side of a threshold, deciding whether and how to bring a new technology across. Yet the organization making that decision is itself becoming different through the encounter, in ways not captured by the measures through which adoption is declared successful.
We are not standing outside this development, directing it from some unaffected position. We are participants in it.
The next time an AI-generated analysis arrives at a meeting, there is a question worth keeping nearby. Not simply whether the analysis is accurate, or how much time it saved.
But whether the people gathered around it have come to understand the matter before them or have merely received a convincing account of it.
For Further Reading:
The questions raised in The Myth of Adoption are also being taken up, in different terms, in current research on AI and organizational life. Two articles from the Boston Consulting Group offer useful perspectives on how AI is changing the work of enterprises.
The Corporate Strategy Function in an AI-First World (2026) explores how AI is reshaping strategic thinking and decision-making.
The AI Adoption Puzzle: Why Usage Is Up But Impact Is Not (2025) examines the gap between AI utilization and meaningful organizational impact.
Read together, these articles invite a question worth carrying into your own organization: As AI changes how work is accomplished, are we attending equally to how understanding and judgment are being formed in us?
