Between 2023 and 2025 the share of Italian companies using at least one artificial intelligence technology more than tripled, rising from 6% to 16.4%.
The figure comes from Istat’s 2026 Annual Report, presented in Rome on 21 May. Behind the average, however, lies a deep divide between large firms and smaller ones.
Adoption at Two Speeds
Large companies, with 250 or more employees, adopt AI in 53% of cases. Medium-sized firms, between 50 and 249 employees, stop at 27%.
Small companies with up to 49 employees reach only 14.2%. The gap therefore widens as company size decreases.
Meanwhile other digital technologies are advancing: data analysis use rose from 26.6% to 42.7% in two years, while ERP systems grew from 33% to almost 50% between 2021 and 2025.
The Real Barrier Is Human Capital
When SMEs that evaluated AI without adopting it are asked about the main obstacle, the answer is one: the lack of internal skills.
Not technology costs, nor uncertainty about returns. More than half of companies with fewer than 250 employees point to skills as the main barrier.
Istat also notes that the average age of staff and seniority in a role are associated with a lower propensity to adopt AI.
Those Who Invest in People Move Faster
Companies with a higher share of graduates record an AI adoption rate 8.8 percentage points higher than those with the lowest share.
For firms with more qualified professionals, such as specialists and technicians, the advantage is 5.2 percentage points. On the general digitalisation index the gap rises to almost 11 points.
Human capital is not only a productivity factor: it is what allows a company to adopt the technologies that, in turn, generate productivity.
Where Italy Falls Behind
The report highlights two structural gaps. Spending on Research and Development remains below 1.5% of GDP, while Germany is at 3.1%, France at 2.2% and Spain has just reached the same Italian threshold.
The spread of the most advanced technologies stays concentrated in large firms: four out of five large companies introduced innovations in the 2020-2022 period.
For SMEs, the path indicated by the data runs through training and the reskilling of existing staff, rather than simply purchasing software.
Original article: wtraining.it
