The real question for businesses is no longer whether to adopt AI, but whether they can run it with clear ownership, oversight, and control.
As AI moves from experimentation to practical use inside companies, the real challenge is less about the model and more about the people and processes that make it work.
Italian companies are being pushed to raise their security maturity, but weak governance, uneven training, incomplete controls, and fragile risk management can make compliance harder than the checklist suggests.
The real bottleneck is not whether generative AI works, but whether companies can govern it, train people for it, and turn experimentation into a controlled operating model.
Manufacturers are moving past pilots and looking for repeatable Industrial AI models that can turn data, processes, and platforms into measurable value.
A study from the ITIR at the University of Pavia puts a sharper question on the table: when artificial intelligence spreads through firms, does it really translate into useful work, or only into visible adoption?
A national push for a European AI gigafactory looks ambitious on paper, but power, financing, and real demand may decide whether the plan becomes infrastructure or just a declaration.