Artificial intelligence, usage among enterprises doubles
By 2025 adoption has risen to 16.4% and exceeds 50% in companies with more than 250 employees. The distance between North and South remains. Italy still 18th in the EU
by C.Fo.
The adoption of artificial intelligence solutions among companies is partially on the rise. This is noted by Istat in its survey 'Enterprises and ICT' updated to 2025. The improvement is evident above all among larger companies, but SMEs are also making initial progress. Large differences remain between the Centre-North and the South.
Overall, the introduction of AI in companies with at least 10 employees doubled from 8.2% in 2024 to 16.4% in 2025 (it was 5.0% in 2023). Adoption grows rapidly as the number of employees increases, from 14% of companies with between 10 and 49 employees to 53% of those with more than 250 employees.
It is precisely the class of larger companies that shows the largest increase (in 2024, adoption was 32.5%), but SMEs are also clearly accelerating (from 7.7% to 15.7%).
As for the territorial breakdown, the North-West has the best performance with 19.3%, ahead of the North-East (17.6%) and the Centre (15.2%). But it drops to 12.2% in the South. There are also visible differences by sector of use. The results are better in those with greater information and creative intensity than in traditional manufacturing sectors. Six sectors, in particular, use AI technologies for at least 25% of the enterprises: IT and other information services; film-video-music products; publishing activities; tlc; professional, scientific and technical activities; travel agencies, tour operators and reservation services. A further breakdown emerges in terms of usage in the various business areas. In marketing and administrative processes, generative AI and text mining techniques emerge; in production processes and logistics, workflow automation techniques; in IT security, predictive machine learning techniques; in research and development, machine learning techniques; and in logistics, solutions for autonomous machine handling.
Another section of the Istat analysis focuses on the limits to the introduction of AI among enterprises that have not yet embarked on this path. Some considerations differ according to company size. Smaller companies highlight, as barriers to use, above all the lack of in-house skills (58.6%), the lack of clarity about the legal consequences of damage caused by the use of AI (47.3%) and the availability of data (45.2%). Larger companies show less concern about costs and the perception that AI may not be useful, while they fear more about data protection and privacy violation issues.
