The article at a glance

  • Artificial intelligence is spreading rapidly across companies and is beginning to change the way work and decision-making processes are organized. 
  • According to research by the University of Pavia, around 60% of workers already use AI tools, but only 17.4% consider them truly central to their work.
  • Governance structures are still limited: only 16.3% of companies have roles dedicated to AI, and 8.5% have specific organizational units. 
  • As many operational activities become automated, human work will increasingly shift toward supervision, goal-setting, and the coordination of intelligent systems.
  • As a result, new professional roles will emerge: AI managers, responsible for integrating artificial intelligence into business processes and guiding the transformation of skills.

In recent years, artificial intelligence has begun to spread rapidly within companies. Its presence is progressively extending into numerous areas of everyday work: from content production to the synthesis of information, from decision support to the automation of certain operational activities.

This evolution is also beginning to change the way companies organize work and make decisions; as a result, new roles are expected to emerge, tasked with governing the integration of artificial intelligence into business processes.

Understanding how companies are dealing with this phase of transformation is one of the aims of the research “Governance and Diffusion of Artificial Intelligence“, conducted by the ITIR research center of the University of Pavia on more than 5,000 professionals from medium-sized and large companies. 

The results of the study, together with the reflections that emerged during a round table attended by Federico Bonelli, CEO of RES IT, help interpret a change that concerns not only the adoption of new technologies, but also the way organizations structure work, skills, and responsibilities.  

AI adoption in Italian companies: what the data tells us

The research data provides a fairly clear picture. Around 60% of workers say they use artificial intelligence tools in carrying out their professional activities. A percentage of this magnitude indicates that the technology has already become a stable part of the everyday working life of many people. At the same time, only 17.4% of the sample considers AI truly important to their work.

This gap between usage and perceived relevance suggests that many applications are still being adopted in an exploratory way. These tools are used for limited activities, often for operational support or personal experimentation. Technology is therefore appearing in work practices, but only rarely, at least for now, does it take on the role of central infrastructure within business processes.  

A further point of interest concerns the ways in which AI spreads within organizations. Spontaneous initiatives by colleagues represent one of the main factors behind the spread of usage practices (43.1% of the sample). Many applications arise from individual initiative, technological curiosity, or the attempt to improve one’s personal productivity. The phenomenon develops through mechanisms of imitation and peer learning, rather than through implementation programs planned by the company.  

In fact, organizational structures dedicated to AI are still limited. Only 16.3% of companies say they have introduced specific roles for AI management, while dedicated organizational units are present in 8.5% of cases and formalized policies concern around 21% of companies.  

Technology is therefore widely present in operational practices, but remains only partially integrated into governance mechanisms. 

The AI paradox: widespread use, limited organizational integration

The research therefore highlights a dynamic that can be described as an organizational paradox. Artificial intelligence is widely present in everyday activities, while its integration into decision-making models and business processes is progressing more slowly.

A significant share of current usage concerns tools that are easily accessible and available. Generative applications, for example, have reached very high levels of diffusion precisely thanks to their ease of use. This encourages rapid adoption, but does not automatically guarantee a meaningful organizational impact.

More established technologies integrated into business systems, by contrast, show more limited diffusion but a better perceived usefulness. These applications are generally used for specific purposes and with more clearly defined operational goals. Their contribution therefore appears more tangible, even if less visible.

The Italian market appears to be in a transitional phase: AI is spreading rapidly among individual workers, while its integration into business processes requires longer timeframes.

A similar dynamic is already visible in technology companies, where artificial intelligence is beginning to reshape work in software development. In organizations made up of hundreds of programmers, experiments with intelligent agents show how a significant share of operational activities can be progressively automated. Code production is increasingly being supported by systems capable of generating it automatically, opening up a phase of transformation in the skills required of developers. 

How should AI use be managed within companies?

Empirical evidence shows that the competitive impact of artificial intelligence grows significantly in companies that have launched more structured integration paths. Companies that invest explicitly in AI projects, introduce dedicated roles, or develop specialized organizational units report much higher levels of perceived value.

These results indicate that the technology expresses its potential above all when accompanied by governance mechanisms. Targeted investment, clear organizational responsibilities, and defined processes make it possible to transform AI from an experimental tool into a strategic lever.

The spread of intelligent systems is also changing the nature of work within organizations. As technologies become more autonomous, purely executive work will tend to decrease. Many operational activities will be delegated to intelligent systems, shifting human contribution increasingly toward functions of direction and supervision.

People will therefore find themselves carrying out activities such as: 

  • defining objectives, 
  • formulating operational instructions, 
  • building the informational context, 
  • verifying the results produced by systems. 

New roles are emerging within companies: AI managers

The evolution just described suggests the progressive spread of a new category of professional roles within companies: AI managers. 

These are roles intended to govern the integration between intelligent systems, business processes, and human activities. Their function concerns coordinating the use of technology, integrating it into operational processes, and defining rules and responsibilities for its use.

According to Bonelli, this transformation could also affect the structure of skills within companies. Many workers who currently carry out operational activities may gradually find themselves coordinating intelligent systems rather than performing the work directly.

This change will therefore require very deep reskilling processes. Professionals used to working directly with tools will need to develop capabilities in supervision, coordination, and the management of complex systems, ultimately handling multiple intelligent agents at the same time, defining objectives, and verifying results. In short, becoming true “managers” of generative tools. 

Within this new family of roles, different specializations may emerge. Some companies are already introducing positions such as Chief AI Officer or AI strategy leads, tasked with defining the overall guidelines for technological development. Other organizations are developing roles dedicated to operational governance, project coordination, or the management of the ethical and regulatory aspects related to the use of artificial intelligence.  

Why AI managers will be needed 

Artificial intelligence is introducing new ways of producing information, supporting decision-making, and organizing work. Systems capable of analyzing data, generating content, and suggesting operational actions are gradually changing the way organizations coordinate their activities.

This evolution requires governance capabilities: AI managers will be tasked with creating the conditions for collaboration between people and intelligent systems to produce value.

As Bonelli points out, this transformation concerns not only companies’ technological efficiency, but also their social responsibility. In an economic fabric made up predominantly of small and medium-sized enterprises, many people have built long-term professional careers within the same company. In these contexts, the challenge will not be to adopt new technologies, but to progressively transform operational workers into professionals capable of guiding complex systems.