The European Banking Authority warned in its June 2026 Risk Assessment Report that rapid development of highly capable artificial intelligence models is creating new cybersecurity risks for banks, adding uncertainty to an operating environment already shaped by geopolitical tensions, volatile energy markets, and growing exposure to non-bank financial institutions. The regulator stated that recent advances in frontier large language models have significantly increased concerns among banks and supervisors, as the latest generation of AI systems has demonstrated increasingly sophisticated capabilities to identify and exploit software vulnerabilities. European banks continue to report strong profitability, resilient asset quality, and capital levels near record highs, yet the EBA's assessment indicates operational resilience is emerging as one of the most important challenges for the industry.
EBA Identifies AI Models as Source of Vulnerability Exploitation Risk
The EBA stated in its June 2026 report that recent advances in frontier large language models have significantly increased concerns among both banks and supervisors. The regulator noted that the latest generation of AI systems has demonstrated increasingly sophisticated capabilities to identify and exploit software vulnerabilities, raising questions about the banking sector's ability to keep pace with rapidly evolving threats.
According to the regulator, recent advances in highly capable AI models are creating concerns that cybercriminals and hostile actors could use these systems to identify weaknesses in software infrastructure more efficiently than before. The report states that frontier AI models are, at a minimum, increasing cyber risk across the banking sector.
The EBA notes that institutions lacking sufficient operational resources may find it difficult to respond quickly enough as AI-driven threats become more sophisticated. That challenge is particularly relevant for smaller institutions that may not possess the same cybersecurity budgets and technical capabilities as larger banking groups.
Operational Risk Category Rises Amid Cyber and Geopolitical Concerns
The EBA identifies operational risk as one of the few major risk categories that continues to move higher. According to the report, cyber risk, data security concerns, and fraud remain key drivers behind the increase. While available data does not yet show a material rise in successful cyberattacks against European banks, supervisors believe the underlying threat environment has deteriorated.
The regulator links part of this increase to geopolitical developments. Geopolitical tensions have heightened concerns around cyber threats, data security incidents, and even physical attacks against critical infrastructure. Financial institutions increasingly operate in an environment where cyber events can originate from criminal organizations, politically motivated groups, hostile states, or opportunistic actors seeking to exploit periods of instability.
The emergence of advanced AI systems introduces the possibility that threat actors can automate vulnerability discovery, accelerate attack development, and reduce the technical expertise previously required to conduct sophisticated operations.
AI Influences Private Credit Markets and Asset Valuations
The EBA report notes that artificial intelligence has already influenced financial market behavior and asset valuations. According to the report, AI-related optimism has contributed to elevated valuations in parts of the equity market, creating potential financial stability concerns if earnings expectations fail to materialize.
Citing research from the Bank for International Settlements, the EBA notes that AI-related projects accounted for more than one-third of private credit deals in 2025, compared with 17% during the previous five-year period. The regulator warns that investors could face losses if valuations linked to AI infrastructure, including data centers, experience sharp corrections.
The EBA notes concerns raised by the Financial Stability Board regarding private credit market transparency, valuation practices, liquidity mismatches, and growing interconnections between private credit funds and banks. The regulator states that AI investments may face risks from construction delays, electricity supply constraints, demand uncertainty, and broader economic conditions.
European Banks Maintain Strong Capital and Profitability Metrics
EU and EEA banks continue to expand their balance sheets despite a more volatile risk environment. Lending to households and non-financial corporates increased by 2.7% during 2025, while exposures to non-bank financial institutions rose significantly and now represent approximately 10% of total assets.
Asset quality remains strong, with non-performing loan ratios near historic lows. Capital ratios remain close to record highs, profitability continues to exceed 10% return on equity, and liquidity metrics remain comfortably above regulatory minimums.
The regulator identifies a growing list of vulnerabilities that banks will need to manage, including geopolitical tensions linked to the Middle East, higher energy prices, sovereign debt pressures, private credit market growth, rising connections with non-bank financial institutions, commercial real estate exposures, and operational risks associated with emerging technologies.
FAQ
What did the EBA warn about AI models in its June 2026 report?
The European Banking Authority warned in its June 2026 Risk Assessment Report that rapid development of highly capable artificial intelligence models is creating new cybersecurity risks for banks. The regulator stated that frontier large language models have demonstrated increasingly sophisticated capabilities to identify and exploit software vulnerabilities, raising concerns about the banking sector's ability to keep pace with rapidly evolving threats.
How much did EU and EEA bank lending grow in 2025?
Lending to households and non-financial corporates by EU and EEA banks increased by 2.7% during 2025. Exposures to non-bank financial institutions rose significantly and now represent approximately 10% of total assets, while profitability continues to exceed 10% return on equity.
