From experimentation to results: AI has come of age. But governance and skills remain the key issue

Artificial intelligence in Italian businesses is entering a new phase: with the era of experimentation and pilot projects now behind us (though not permanently), many organisations are slowly beginning to see tangible results in terms of efficiency, productivity and return on investment. This is essentially the picture that emerges from the “The SAP Value of AI 2026” study, carried out by the German software giant in collaboration with Oxford Economics and previewed during the “Meet & Grow” at Sole 24 Ore Business School – a forum that brings together CEOs, managers, students and professionals to discuss innovation and corporate strategic decisions. The special guest at yesterday’s session – which also covered open systems and human capital – was Carla Masperi, managing director of SAP Italia. “We are integrating AI into business processes not merely as a matter of automation – to boost and improve individual productivity – but also as a matter of decision-making, because agent technology is changing the way decisions are made, based on real-time information. “AI,” the manager continued, “must become part of our daily routine, and we must encourage people’s ability to keep pace with the speed of technology in order to reap all the benefits.” Alice Acciarri, Managing Director of Sole 24 Ore Business School, also spoke about personal development, emphasising that “the value of training increasingly lies in the ability to create opportunities for interaction between the academic world and business. “This is why we are delighted to host organisations such as SAP, which can offer our community practical insights, skills and perspectives that are useful for interpreting change and tackling the challenges of the future,” she added.

Masperi then shared with the audience some findings from the study, which involved over 2,300 managers in 13 countries – 200 of whom were from medium-sized and large companies in Italia. The ecosystem as a whole views artificial intelligence (including agent-based AI) with growing confidence and with the realisation that value creation will increasingly depend on the ability to integrate technology, processes and governance. This is despite the fact that significant challenges remain in terms of data quality, training and governance. In terms of spending, Italian companies plan to allocate an average budget of 18.4 million dollars to AI projects this year – a figure lower than the global average (28 million dollars) but set to grow by 45 per cent over the next two years.

This increase in spending is accompanied by ever-higher expectations regarding financial returns, with an estimated average ROI of 20% in 2026 (the equivalent of approximately $4.9 million) and set to rise to 38% within the following two years (reaching $12.2 million). According to experts, this figure is highly indicative of AI’s transformation from an emerging technology into a strategic lever for competitiveness, which the CEO of SAP has described as follows: “Over the past year, Italia has moved from the experimental phase to the operational phase, and this is beginning to yield concrete results. But there is still a long way to go, because AI without context – be it in terms of processes, data or governance – generates activity that yields no results at best, and creates new risks at worst.”

The level of technology adoption shows that a quarter of business activities (26 per cent, to be precise) are already supported by artificial intelligence, with this figure expected to reach 44 per cent by 2028. Does this therefore pave the way for the maximum use of generative tools and agents? Not exactly. The path forward remains uneven, as evidenced by the fact that 20 per cent of companies report having adopted a strategic approach to AI investment, whilst the largest proportion – 36 per cent – continue to proceed through fragmented and not fully integrated initiatives.

Organisational and cultural factors are the main obstacles, and one of these is the lack of dedicated staff: in fact, fewer than half of the companies surveyed have a manager dedicated to artificial intelligence on their staff, whilst only 35 per cent define specific performance indicators to measure its impact and just 37 per cent offer structured training programmes on the benefits and risks of the technology. Despite this, 73% of organisations say they are satisfied with the financial return achieved to date, whilst acknowledging that the technology’s potential is still far from being fully realised.