Reconstruction
reconstruction
Establishing what actually happens in a workflow from sources that are not people's recollection, then setting that against what is declared. The object produced is not the list, it is the distance between the two, and the evidence status of every element of that distance.
Process mining already does this for software processes: it reads event logs and reconstructs how work really runs. Its limit is its input. Shadow AI leaves no exploitable log, because the work passes through a browser tab, a personal account, a copied and pasted paragraph.
Reconstruction is therefore process mining applied where there are no logs. The inputs become the documents an organisation can produce, targeted interviews, and the cross-checking of the two. A finding is only established once a second source of a different type confirms it, which is what replaces the log as the guarantee.
What it means for an SME running AI
The part of your AI use that leaves traces is already measurable by tools you can buy. The part that does not is the part that has been growing for eighteen months.
Sources
- Par différence avec le process mining, discipline établie qui reconstruit le déroulement réel du travail à partir des journaux d'événements des systèmes.
This is what the survey measures See how it is measured →
Or answer six questions on your own uses Take the check →