The shift in perspective
Making an artificial intelligence tool available to employees is not enough to generate results. This is the thesis at the center of an analysis published by Agenda Digitale, which cites research released in 2026 by Gartner, the Osservatori del Politecnico di Milano and Deloitte, together with testimony from Emiliano Papadopoulos, chief executive officer of Allnet.Italia. The source notes that AI adoption in companies is growing, but that translating it into concrete benefits requires three conditions: reliable data, redesigned processes and clear responsibilities. The starting point, therefore, is not the choice of software but the organization of the work that must accommodate it.
According to the source, AI helps reduce the gap between the speed at which an individual completes a task and the actual result for the company. This is an important distinction: individual speed does not automatically translate into value if the context in which the work is produced has not been redesigned. Hence the recommendation to invert the usual order of priorities, addressing processes first and tools second.
Governance, data and security
The source identifies three fronts on which to focus attention when an AI tool is made available to the entire company workforce. The first is governance: oversight is needed over how information circulates. The second concerns data protection. The third is security. Two organizational questions are added to these: who checks the errors produced, and what happens to the work freed up by automation.
These are issues that do not depend on the technical capabilities of the system adopted, but on the rules and roles built around it. The source does not provide data on the number of companies involved or details on the results of the cited research; however, it indicates the direction of change, placing it in the redesign of workflows rather than in the purchase of new applications.
What remains to be verified
The available material is limited to indicating the sources of the research and their year, without reporting their specific contents. It is therefore not possible to establish which sectors are more advanced, which obstacles have been measured most frequently, or which concrete results have been observed in the cited companies. Papadopoulos's testimony is mentioned as part of the picture, but no direct statements by him are reported.
The question of employment consequences also remains open: the source poses it as a question to be governed, without offering elements to assess it. The point that emerges clearly is methodological: the order of priority between processes and tools. Verifying how companies are actually applying this principle would require the data from the cited research, which are not present in the material provided.
