The European Central Bank published a study of artificial intelligence adoption and investment based on 2025 Survey on the Access to Finance of Enterprises data covering more than 6,000 firms in 12 euro area countries. About 70% of firms reported some AI use, but only 7% described their use as significant. Adoption exceeded 80% in the Netherlands, Finland and Austria but was below 65% in Italy and Ireland. Larger, younger and technology-intensive firms were more likely to adopt AI, although significant use remained limited across all firm sizes. Firms primarily use AI to improve business processes, while skills shortages, data privacy and ethical concerns, and incompatibility with existing systems are the main barriers to wider use. AI adoption is financed mainly through internal funds, supplemented by grants and subsidised bank loans. Significant AI users reported stronger expectations for turnover, fixed investment and selling prices. The study found no statistically significant economy-wide productivity difference between users and non-users, although positive associations were evident in information and communications technology and research and development. It also found no evidence of workforce reductions or a broad relationship between current AI use and inflation expectations. Firms planned to allocate an average 9% of total investment to AI over the following year, rising to 20% among significant users and falling to 4% among non-users. A survey experiment found that firms underestimated competitors’ AI investment by an average 18 percentage points. Providing accurate information raised expectations of competitors’ future adoption by about 4 percentage points and firms’ own planned AI investment rates by nearly 2 percentage points.
European Central Bank2026-07-28
European Central Bank study finds widespread but shallow AI adoption across euro area firms
A European Central Bank study found that about 70% of euro area firms use AI, but only 7% report significant adoption, with skills, privacy concerns and system incompatibility constraining wider use. AI investment relies mainly on internal funds, and intensive users report stronger turnover and investment expectations without evidence of workforce reductions. Firms substantially underestimate competitors’ adoption, while correcting those perceptions increases their own planned AI investment.