The DIKW pyramid is a mental‑model that orders four increasingly abstract concepts – Data → Information → Knowledge → Wisdom – as successive layers of meaning built on one another [1][2][3].
| Layer | What it is | How it is created |
|---|---|---|
| Data | Raw, unprocessed facts, numbers, symbols or signals that have no intrinsic meaning on their own. | Collected from sensors, surveys, logs, etc.; it is the “raw material” for all further processing [1][2]. |
| Information | Data that has been given context, structure or relevance, turning it into something that can be understood. | Organization, classification, tagging or linking of data so that a pattern or relationship becomes apparent [1][2]. |
| Knowledge | Synthesized information that is combined, compared and interpreted over time, producing insights about how things relate and how they work. | Integration of multiple pieces of information, applying expertise, experience and mental models to draw conclusions [1][2][3]. |
| Wisdom | Judicious, value‑laden judgment about when and why to act on knowledge; it incorporates ethical, cultural and long‑term perspectives. | Evaluation of knowledge against goals, principles and societal impacts; it guides decision‑making and behavior [1][2][3]. |
The pyramid illustrates a progressive enrichment: raw data become useful information only when placed in a context, information becomes knowledge when it is organized and understood, and knowledge becomes wisdom when it is tempered by values and foresight. In practice, the model is used to identify where problems lie (e.g., “knowledge security” versus “data security”) and to avoid trying to solve a higher‑level issue (such as wisdom) by applying controls only at the lower data layer, which can “squeeze” the whole pyramid and produce ineffective or harmful AI outcomes [1].
Because each step adds a new abstraction layer, governance frameworks (for AI, data policy, or diplomatic knowledge management) must design controls appropriate to the specific layer they aim to protect [1][3]. This way, policymakers can move beyond treating all digital assets as mere “data” and instead recognise the distinct challenges of protecting information, fostering knowledge sharing, and nurturing the wisdom needed for responsible global governance.