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

Other formats

• Markdown version
• Structured JSON version
• Bundle manifest

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