## How Central Banks Can Contain Financial Stability Risks as AI Accelerates Change _IMF Blog, July 23, 2026_ ## Source details **Canonical URL:** [How Central Banks Can Contain Financial Stability Risks as AI Accelerates Change](https://www.imf.org/en/blogs/articles/2026/07/23/how-central-banks-can-contain-financial-stability-risks-as-ai-accelerates-change) ## Other formats - [Markdown version](/en/blogs/articles/2026/07/23/how-central-banks-can-contain-financial-stability-risks-as-ai-accelerates-change/index.md) - [Structured JSON version](/en/blogs/articles/2026/07/23/how-central-banks-can-contain-financial-stability-risks-as-ai-accelerates-change/index.json) - [Bundle manifest](/en/blogs/articles/2026/07/23/how-central-banks-can-contain-financial-stability-risks-as-ai-accelerates-change/bundle-manifest.json) ## Bibliographic details - Authors: Tobias Adrian - Published: July 23, 2026 --- ### Three priorities - Strengthen oversight and governance of AI‑driven trading, lending, and supervisory technology (SupTech). - Improve visibility into AI use, dependencies, and asset correlation risks due to synchronized trading strategies. - Deepen international cooperation on operational resilience and cyber defense. ### How AI changes market dynamics - AI compresses time and distance in finance, moving trading, credit decisions, and supervisory analytics increasingly into real time. - Trading and lending - AI is increasingly embedded in trading and investment across major capital markets: machine‑learning models generate high‑frequency signals; generative AI parses earnings calls, regulatory filings, and economic news in real time. - Under normal conditions, AI‑driven execution can improve liquidity, lower transaction costs, and accelerate price discovery. - In credit markets, AI‑supported models in consumer and small‑business lending strengthen fraud detection and broaden data used for risk assessment. - During stress, AI can amplify market moves: some AI‑based funds rebalance much faster than traditional strategies, amplifying swings when many models respond to similar signals. - Future flash crashes may arise less from coding errors and more from many AI systems reacting in parallel to the same information. - Opacity of AI models makes it harder for authorities to explain behavior under stress and to detect emerging risks. - Policy implications for markets - Central banks and supervisors must monitor AI‑driven strategies more closely, map correlation risks, and ensure stress testing captures the speed, scale, and interactions of AI‑based decision‑making. - Enhanced monitoring and better data on AI adoption, model dependencies, and market exposures should complement traditional capital and liquidity buffers. - Greater transparency about model use can help identify systemic vulnerabilities. - Well‑governed AI could mitigate human biases and diversify decision‑making, contingent on strong safeguards around model risk and transparency. ### Infrastructure, concentration, and operational resilience - AI is transforming operational cores: back‑office automation, real‑time workflows, and AI use in financial market infrastructures (exchanges, clearinghouses, payment systems) for monitoring and anomaly detection. - Main systemic risk: concentration of critical services and shared dependencies on a small number of cloud, data, or model providers. - Disruptions at a critical provider—technical, cyber‑related, or geopolitical—could affect many institutions simultaneously. - Regulatory responses observed - Several authorities, including the European Central Bank and the Bank of England, have expanded operational‑resilience frameworks to explicitly cover critical third‑party service providers, including AI and cloud vendors. - Policy recommendations for infrastructure - Conduct system‑wide mapping of AI‑related dependencies. - Set minimum resilience standards for key providers. - Develop contingency planning for correlated outages. - Avoid policy choices that implicitly favor a small set of approved AI providers or encourage firm convergence that increases common dependencies and correlated failures. - Balance benefits of known suppliers against systemic risks from excessive concentration and conformity. ### Risk management and supervision (SupTech) - SupTech applications - Central banks and supervisors use machine‑learning and natural‑language processing in surveillance and to spot anomalies and emerging risks (examples: authorities in France, Portugal, Germany, and Japan; Federal Reserve, ECB, and Bank of Canada). - Benefits and risks - SupTech can improve timeliness, coverage, and analytical depth. - Governance challenges include scarce specialized expertise, especially in emerging markets; model risk; and over‑reliance on automated outputs creating blind spots during stress. - Supervisory principles and actions - Adopt the principle that AI should augment supervisory judgment, not replace it. - Implement robust governance, explainability requirements, and human oversight for SupTech. - Invest in supervisory capacity and consider drawing on specialized external institutions to assess complex AI systems. ### AI‑enhanced cyber threats and collective defense - Generative AI increases speed, scale, and sophistication of cyberattacks: more convincing phishing, adaptive fraud, and shortened windows between discovery and exploitation. - Cyber risks have macro‑financial implications as AI enhances attacker capabilities, leaving institutions less time to detect and respond. - Observed initiatives - A Bank for International Settlements survey finds most central banks are adopting or planning to adopt generative AI for threat detection and response, while recognizing the same tools strengthen attackers. - Authorities in Japan have worked with major institutions to assess preparedness for AI‑driven cyber risks. - Work by the Group of Seven highlights AI‑enabled threats as shared vulnerabilities requiring coordinated responses. - Policy recommendations - Strengthen expectations for cyber resilience. - Conduct system‑wide exercises that include AI‑enabled scenarios. - Improve information‑sharing on threats and defenses. - Invest in defensive AI within clear guardrails. ### Overarching message and the IMF’s role - AI is a financial‑stability issue spanning markets, institutions, infrastructures, and supervision; stability will depend on institutions, incentives, and safeguards governing AI use rather than on any single model. - Early and collective policy action can enable AI to reinforce global financial resilience; inaction could make future instability faster, more correlated, and harder to manage. - The IMF can assist countries to identify emerging vulnerabilities, share experiences, and develop policy frameworks through surveillance, financial‑sector assessments, and capacity development. *Tobias Adrian — July 23, 2026* --- _Source: https://www.imf.org/en/blogs/articles/2026/07/23/how-central-banks-can-contain-financial-stability-risks-as-ai-accelerates-change_