ECB survey: 72% of firms planning AI investment will fund it themselves

Serguey Asael Shinder02 October 2026, 17:482 min read

Euro area companies expect to fund their investment in artificial intelligence overwhelmingly from their own resources, according to a post on the ECB Blog by Annalisa Ferrando, Sara Lamboglia, Judit Rariga and Maurice Schmidt. It draws on the latest Survey on the Access to Finance of Enterprises, which asked about 5,000 firms about their AI plans for the next 12 months.

Asked what they plan to invest in, 49% of firms named AI technologies and tools, 46% employee training, 40% data infrastructure and 12% hiring AI specialists; 38% chose none of these. Larger firms are more likely to invest in every category, but the pattern is the same across sizes.

On financing, 72% of firms planning AI investment expect to use internal funds such as cash flow or retained earnings. Bank loans, grants and leasing are each mentioned by 16%, equity and venture capital by 6% and debt securities by 1%, while 18% have not yet chosen a source. Most firms expect to use a single instrument, usually internal funds; where they combine sources, internal funds are typically paired with an external one.

The authors find that the type of investment decides whether external money is used. Tangible assets such as hardware and data infrastructure can be pledged as collateral, and comparing firms in the same country, industry and size class, plans to invest in AI tools or in data and infrastructure are associated with a 16 percentage point higher probability of combining internal and external finance. Hiring AI specialists raises it by 9 points, and employee training has no statistically significant effect, because these investments are intangible and hard to borrow against.

Why it matters

If AI investment is financed mostly from cash flow, the ECB's interest rate decisions reach it only indirectly, and firms without strong earnings fall behind. The finding that skills and training are the hardest part to finance points to where public support or new lending models would make the most difference.