Italy’s artificial intelligence market reached €1.8 billion in 2025, growing 50% year-on-year and reaching a value almost nine times higher than in 2018. The generative component already accounts for 46% of the total, AI skills appear in 76% of postings for qualified profiles, and job listings requiring them have grown 93% in a year. Yet the average adoption rate among Italian companies stands at just 8.2%, compared to a European average of 13.5%: the country is growing fast, but starting from further behind.
These are the findings of “Italy 2030. Scenarios for Innovation between AI, Business and Politics,” the report from Rome Business School edited by Valentino Megale, Program Director of RBS’s International Master in Artificial Intelligence, and Valerio Mancini, Director of RBS’s Research and Outreach Center. “Italy has the assets to build niche AI leadership that no other European country can replicate. Italian industrial districts hold unique operational data that cannot be purchased elsewhere: whoever transforms it into vertical AI models will lead productivity in their sector at the European level,” says Valentino Megale.
Compared to its European partners, the picture is clear: Italy is today a rapidly expanding market but still a follower. France, Germany, the Netherlands, and the United Kingdom have consolidated national strategies and innovation hubs that are years ahead of Italy’s trajectory. The European Commission’s European Innovation Scoreboard 2025 classifies the country among the Moderate Innovators, with performance at 93% of the EU average — a contestable position, with concrete room for catching up.
Adoption figures show where the delay is concentrated. 71% of large Italian companies have launched at least one AI project, and 84% have already purchased licenses for at least one Generative AI tool, a 31% jump compared to the previous year (Osservatorio Artificial Intelligence, Politecnico di Milano, 2025). Among SMEs, the adoption rate falls to 8%. The average rate among Italian companies stands at 8.2%, against a European average of 13.5% (ISTAT, 2024). Since the country’s production structure is made up overwhelmingly of small and medium enterprises, the technology gap between large groups and SMEs risks becoming permanent.
Human capital also weighs on the delay: Italy has a share of the population with tertiary education and a number of ICT specialists both below the European average. In 2025, Italy’s venture capital ecosystem recorded 419 deals worth over €2.3 billion (Venture Capital Monitor, VeM, LIUC/AIFI), compared to the $194 billion concentrated by the United States in the AI market alone, equal to 75% of the global value of deals. Italian startups are being founded, but then struggle to scale.
Large groups in the telco and media, energy, finance, and insurance sectors have built centralized MLOps platforms and agentic systems capable of automating complex workflows. In large companies, the transformation is already visible in the workplace: in 2025, nearly one in two workers used AI tools, and four in ten say AI enables them to carry out activities they otherwise couldn’t (POLIMI, 2025). The challenge for the next cycle is to use the industrial weight already gained to pull the entire supply chain, including supplying SMEs, into the digital transformation.
The most emblematic case of how manufacturing, design, and AI can converge comes from eyewear: in 2025 EssilorLuxottica sold over 7 million Ray-Ban and Oakley smart glasses, more than triple the previous year, with the category contributing to over a third of the group’s growth in some quarters. The same logic drives the Stellantis-Accenture-NVIDIA partnership, announced in May 2026 for AI-enabled digital twins across the group’s global factories. “Italian manufacturing is the natural ground on which to graft digital twins, simulation, and physical AI,” the authors write. “Turning traditional production excellence into a data-driven advantage is exactly what Italy does best; now it needs to be done at a systemic level, not just within large groups.”
In September 2025, Italy became the first EU country to adopt a comprehensive national law aligned with the European AI Act (Law 132/2025), entrusting governance to AgID and the National Cybersecurity Agency. Whoever sets the standard first gains a structural advantage in markets that go on to adopt that model — the so-called “Brussels Effect,” already seen with GDPR.
On the infrastructure front, transformation is underway: Italy has 209 active data centers (as of October 2025), with TIM Enterprise announcing a €1 billion three-year plan and Oracle opening its second Italian cloud region in Turin. In 2025 the European Commission launched a €180 million tender for sovereign cloud services, and Italy is entering with an asset already operational: the Polo Strategico Nazionale, which hosts critical data for central government administrations. The authors suggest strengthening this model by looking at Estonia’s Data Embassy experience — data centers physically hosted abroad but under the full legal control of the country of origin — a viable solution for digital healthcare and critical infrastructure. Rounding out the picture, three of the digital euro infrastructure’s data centers will be located in Italy, with at least one in Rome: a positioning with concrete effects on institutional prestige, talent attraction, and influence at European tables on cloud, data, and digital financial services.
To move closer to niche leadership, the report puts forward a concrete proposal: a “District Dataspace,” a national initiative that would pool — under cooperative governance and responsible AI rules — the industrial data of sectors where Italy already holds an advantage over any competitor: fashion and luxury, precision mechanics, agri-food, pharmaceuticals, and nautical. This data would become the training base for Italian vertical AI models, hosted on the existing sovereign cloud infrastructure and funded by a public-private co-investment fund capable of gradually taking over from the PNRR (National Recovery and Resilience Plan). With 419 venture capital deals and €2.3 billion invested in 2025, the ecosystem has the critical mass to sustain this transition, provided capital concentrates on the right verticals.
“Italy’s production fragmentation, historically its limitation, becomes an asset here,” says Megale. “Each district’s data is unique, non-replicable, unavailable elsewhere. Whoever controls supply-chain data drives productivity in that sector.”
By 2030, Italy can follow three paths. It could fall behind, with AI adoption concentrated in large groups and SMEs left out, eventually becoming a net importer of technology produced elsewhere. It could gradually converge toward the European average, closing the gap without ever becoming a benchmark. Or it could build niche leadership in verticals where no other European country can compete: advanced manufacturing, agri-food, life sciences, space economy.
The third path is achievable, but depends on three variables: continued investment after PNRR funds run out, the speed of skills diffusion beyond centers of excellence, and the ability to close the size gap between large groups and SMEs. The gap with those who have already committed is significant: over the past decade the US has cumulatively invested over $470 billion in AI, compared to $50 billion for the entire EU (OECD, 2025). Without cohesive choices, the second scenario remains the most likely outcome.
“Italy doesn’t need to chase every technological frontier,” warns Mancini. “It needs to concentrate resources where it has a real, defensible advantage. Scenario 3 is achievable, but it requires selective, cohesive choices — not dispersion.”