Small banks face strategic risks as AI reshapes global financial competition

The rise of artificial intelligence is transforming financial services, but its effects on small and medium-sized financial institutions remain underexplored and uneven. A new peer-reviewed study, Meta-Analysis of Artificial Intelligence’s Influence on Competitive Dynamics for Small- and Medium-Sized Financial Institutions, published in Analytics, warns that while academic literature broadly frames AI as an opportunity, it fails to capture how size-specific factors shape outcomes.

The authors argue that the lack of clear, shared language and frameworks leaves smaller institutions without tailored guidance at a time when competitive stakes are rising.

Is AI adoption beneficial for smaller institutions?

The research analyzed 160 publications on AI adoption in finance from 2020 to 2024 using a combination of systematic review, sentiment bibliometrics, and network analysis. Across the literature, the prevailing sentiment toward AI adoption is strongly positive, with marked optimism following the release of generative models like ChatGPT. Expressions of anticipation and future-oriented enthusiasm dominate, suggesting that scholars and practitioners largely view AI as a catalyst for growth, efficiency, and competitive advantage.

Yet the study cautions that enthusiasm may mask real challenges for smaller institutions. While nearly 40 percent of small and medium enterprises report using some form of AI and 26 percent specifically experiment with generative AI, only 8 percent reach transformative integration. At the same time, industry reports suggest that AI project failure rates exceed 80 percent, pointing to a gap between ambition and execution.

For small and medium-sized institutions, the risks are amplified by limited financial and human resources. Unlike large banks with vast datasets and in-house expertise, smaller firms must often rely on external vendors, exposing them to dependency and lock-in. AI’s capital-intensive nature, combined with the need for robust security and compliance, means that adoption is rarely straightforward.

Why does the literature lack clarity on competitive effects?

One of the key findings of the study is the fragmented nature of the academic discourse. Using network analysis of keywords, the authors show that scholarship on AI in finance is highly clustered, with little connectivity across subtopics. The field has not converged on a shared conceptual language or unified frameworks to explain how AI reshapes competition.

This fragmentation leaves blind spots. Popular models like the Technology–Organization–Environment (TOE) framework and the Technology Acceptance Model (TAM) are often applied, but they tend to focus on how adoption happens rather than whether it should happen, especially in contexts where smaller firms may not be ready. The assumption that adoption is inherently beneficial risks overlooking critical size-specific challenges.

The lack of systematic attention to firm size means that academic studies often extrapolate from large financial institutions to the sector as a whole. As a result, the unique pressures on smaller firms, including their constrained capital, data scarcity, and talent shortages, are under-analyzed. Without clearer models, policy and strategy guidance for SMEs remains generic and potentially misleading.

What risks and opportunities do SMEs face in an AI-friven market?

AI adoption is not a level playing field. Large financial institutions gain advantages from scale, with extensive data pipelines, feedback loops, and the ability to invest heavily in model development. These advantages reinforce market concentration, potentially widening the gap between incumbents and smaller players.

For SMEs, opportunities still exist, particularly in niche markets where agility and personalization matter. Cloud-based tools, open-source frameworks, and synthetic data offer pathways to experiment without the same scale of resources as major banks. By targeting specialized lending products or community-focused services, smaller institutions can differentiate themselves even in an AI-saturated environment.

However, the risks remain significant. Over-reliance on external vendors could tether SMEs to black-box systems that they cannot fully audit or adapt, creating both strategic and regulatory vulnerabilities. Privacy and security concerns loom large, with 72 percent of SMEs reported to operate with weak safeguards. The danger of skill atrophy also emerges: as automated decision support becomes more common, smaller teams may lose expertise in core analytical functions.

The authors note systemic risks as well. If many firms rely on similar AI models trained on overlapping datasets, herding behaviors could intensify, amplifying volatility in financial markets. This risk channel remains under-examined in current scholarship, adding urgency to the call for more size-sensitive research.

A call for size-specific guidance

The findings highlight an urgent need for more targeted frameworks that consider the realities of smaller financial institutions. Rather than assuming adoption is inevitable, the authors argue that firms must first assess strategic urgency: not only how to adopt AI, but whether adoption aligns with their operational capacity, data maturity, and risk tolerance.

For practitioners, this means pursuing hybrid strategies that balance vendor solutions with internal data pipelines, reducing dependency and enabling adaptability. For policymakers, it underscores the importance of differentiated regulation and guidance that recognize the uneven playing field between large and small actors.

The study also identifies future research priorities, including linking AI investment levels with profitability and market share, and conducting cross-country analyses to capture uneven adoption patterns. Such work could help close the gap between optimism in the literature and the practical realities faced by SMEs.