Data, skills and governance: Constraints to growth
As mentioned earlier, data quality remains one of the key challenges in the development and seamless integration of AI into business processes. While 67% of businesses say they are prepared in this regard, the impact is not only visible but easy to imagine, with 74% reporting issues with the reliability and quality of the information they use. In fact, 77% of companies surveyed said they experienced delays, rework, or backlogs due to output produced by models that turned out to be inaccurate or incomplete.
However, another important challenge is related to skills. 80% of managers believe that professional development programs are not evolving at the same pace as artificial intelligence tools, and the phenomenon of so-called “shadow AI” or unsupervised use of AI applications by employees (reported by more than two-thirds of organizations) exacerbates the apparent shortfall.
Finally, the issue of governance appears to be even more important. Only 1 in 10 companies believe they are truly prepared to manage and oversee artificial intelligence in terms of skills, processes, and organizational models, a critical issue that will become increasingly important as autonomous agents become more prevalent.
Agent-based AI challenges
Many observers see AI as the next frontier in intelligent automation, with three out of four Italian companies believing that AI agents can have a moderate to very large impact on their organizations. Looking at the numbers, it is estimated that the value created by these tools could generate a return on investment of approximately $13.7 million within two years in Italy. However, the level of enterprise readiness is currently limited, with only 1% of enterprises declaring that they are fully prepared to deploy and manage advanced agent-based systems. Specifically, 40% of enterprises do not yet have a “human” process in place to monitor agent-based workflows, 25% lack a dedicated access control and authorization system, and less than half maintain an up-to-date registry of agents used within their organization. It’s no surprise, then, that nearly two-thirds of companies believe they’re deploying AI agents faster than they can manage them.
“Italian companies must understand that artificial intelligence often offers value that is harder to measure than expected and comes with the risk of evolving at a pace that is too rapid for most organizations to keep up with. AI governance is becoming the most important challenge that many companies do not yet realize they face,” Masperi reflects. And it is precisely this balance of innovation, control, and expertise that will most likely determine the next stage of artificial intelligence adoption. The ultimate goal, especially for SAP, is a model in which processes, people, and intelligent agents work in an integrated manner, the so-called autonomous enterprise. Achieving this requires technological investment and, above all, organizational and cultural change. The CEO of SAP Italy concluded: “Creating real value from AI is not easy, as it requires an entirely new approach. All companies in Italy, large and small, need to integrate AI with the data and operations that drive their business, while ensuring that the technology is subject to the necessary governance to deliver reliable results.”
