AI Is Rewriting Europe's Banking Rules as Smaller Firms Challenge Market Leaders, ECB Study Finds

A European Central Bank study finds that AI exposure is linked to 5% lower assessed default risk, while smaller businesses gain a competitive edge over established market leaders. The findings highlight the need for smarter lending policies, workforce reskilling and stronger financial supervision to ensure AI-driven growth benefits businesses without increasing financial risks.

Artificial intelligence is changing how European banks assess business risks, distribute loans and evaluate which companies are likely to succeed. A new study published by the European Central Bank (ECB) in October 2026 suggests that generative AI could improve financing opportunities for businesses while weakening the traditional advantages of established market leaders. The research was conducted by Bruno Buchetti, Christian Eufinger, Luca X. Lin and Ixart Miquel-Flores. Its findings offer important lessons for governments, financial regulators, international development institutions and businesses navigating the economic impact of artificial intelligence.

The researchers analysed approximately 6.8 million firm-bank-quarter observations involving private companies in Germany, France, Italy and Spain between 2021 and 2024. Using banking information from AnaCredit and company financial records from Orbis, they examined how lenders changed their assessments of business creditworthiness following ChatGPT's public release in November 2022. The findings show that expectations about AI can influence financing decisions even before companies demonstrate measurable productivity improvements.

AI Is Making Banks More Optimistic, but Not Every Business Benefits Equally

European banks increasingly view artificial intelligence as an opportunity to improve productivity, reduce operating costs and strengthen business performance. The study found that a one-standard-deviation increase in public attention to AI was associated with approximately a 5% reduction in assessed default risk for companies operating in the most AI-exposed industries, relative to other borrowers.

Industries such as computer programming, information services, legal and accounting activities, publishing and scientific research showed particularly high exposure to AI. These sectors could benefit from automation, faster information processing and improved decision-making.

However, labour-intensive businesses face greater uncertainty. Companies with relatively low labour intensity experienced an approximately 7% reduction in assessed default risk, compared with only 3% among labour-intensive firms.

For policymakers, this difference highlights the need for workforce training, digital skills development and financial support for businesses adapting to technological change. Without such measures, AI-driven productivity improvements could be accompanied by workforce disruption and uneven economic benefits.

AI Could Give Smaller Businesses a Competitive Edge Over Market Leaders

One of the study's most important findings is that AI may reduce the competitive advantages traditionally enjoyed by large, established companies.

In highly AI-exposed markets, a one-standard-deviation increase in AI attention was associated with approximately 8% higher assessed default risk for market leaders relative to the smallest competitors, compared with changes in less-exposed markets. For companies in the top 10% of their markets, the relative increase approached 10%.

This does not mean leading companies necessarily became financially weaker. Instead, banks viewed smaller competitors more favourably, suggesting that AI could make it easier for emerging businesses to challenge established players.

The changing assessments were accompanied by shifts in financing. Market leaders experienced approximately 3.8% less drawn credit and 2.4% less committed credit relative to smaller competitors, compared with corresponding changes in less-exposed markets.

Their probability of losing an existing banking relationship increased by 1.2 percentage points, while the probability of attracting a new lender declined by 0.6 percentage points.

For governments and development partners, these findings suggest opportunities to strengthen entrepreneurship, improve access to finance and encourage competition through targeted AI adoption programmes.

Banking Expertise Matters as AI Creates New Financial Risks

The research found that banks with strong industry knowledge were considerably more optimistic about AI-exposed borrowers than other lenders.

Banks with at least 10% of their corporate lending portfolios concentrated in a borrower's industry assigned default-risk estimates approximately 11% to 16% lower than non-specialist lenders assessing the same AI-exposed businesses, relative to other borrowers.

Interestingly, longstanding relationships between banks and companies did not produce similar differences, suggesting that industry expertise is particularly important when evaluating technological disruption.

The study also found that market leaders' traditional advantage in avoiding repayment difficulties narrowed by approximately 0.5 to 0.7 percentage points in highly AI-exposed markets, against an average adverse-credit-event rate of 8.2%.

For financial regulators, these findings highlight the importance of monitoring lending concentrations and ensuring that optimism about AI does not encourage excessive financial risk.

Governments and Development Partners Need an AI-Ready Financing Strategy

The findings suggest that artificial intelligence could reshape economic development by changing both business productivity and access to credit.

Governments should prioritise affordable digital infrastructure, workforce retraining, support for small businesses and stronger financial supervision. International development partners could use these lessons to design financing programmes that encourage responsible AI adoption while strengthening banks' ability to assess technological risks.

For private-sector stakeholders, smaller businesses may gain opportunities to reduce costs and compete more effectively, while established companies will need to modernise operations, invest in employee skills and adapt existing business models.

However, the researchers caution that the study measures banks' expectations rather than proving that AI adoption directly improves financial performance. It also cannot fully separate lending decisions from changes in borrowers' demand for credit.

Future policies should therefore focus on measuring actual productivity gains, employment effects, business competitiveness and loan repayment outcomes.

The central message is that AI could transform not only how businesses operate but also which companies receive financing and gain competitive advantages. Governments, development institutions and financial regulators must ensure that this transformation supports innovation, inclusive economic growth and financial stability without creating new risks for businesses and banking systems.