Invisible crisis: AI-powered cyber-attacks can endanger financial stability

The IMF warns that AI-powered cyberattacks could threaten global financial stability, exposing vulnerabilities in digital payment systems and banking infrastructure

An analysis published recently by the International Monetary Fund (IMF) warns that artificial intelligence (AI) is increasing risks to financial stability by making cyberattacks faster, cheaper and more sophisticated. Advanced models can now identify vulnerabilities in software, payment systems, and Cloud infrastructure at machine speed. Anthropic’s “Claude Mythos” model reportedly demonstrated the ability to exploit weaknesses across major operating systems and browsers. Such developments dramatically reduce the time and cost required to launch attacks, widening the gap between attackers and defenders. This changes the nature of cyber risk itself. Modern finance is deeply interconnected through common software, Cloud service providers, payment networks, and digital infrastructure. A single vulnerability in a widely used system could ripple across multiple institutions simultaneously. Unlike traditional banking crises triggered by bad loans or asset bubbles, AI-driven cyber shocks could emerge suddenly through payment failures, frozen liquidity, or loss of public trust in financial intermediaries. The IMF warns that such “correlated failures” could create funding strains, solvency concerns, and broader market disruption.

 

The threat is serious for India as well. Over the last decade, India has built one of the world’s largest digital public infrastructures. The annual volumes of transactions on Unified Payments Interface (UPI) expanded from just 20 million in 2016-17 to more than 24 billion in 2025-26, while fintech adoption has expanded rapidly across banking, insurance and lending. This digital transformation has increased efficiency and financial inclusion, but it has also expanded the attack surface. A coordinated cyberattack on payment systems, Cloud infrastructure, or banking networks could disrupt economic activity at scale and trigger panic among consumers and investors. Recognising this, the Reserve Bank of India last year had proposed the Framework for Responsible and Ethical Enablement of Artificial Intelligence (FREE-AI), which seeks to ensure responsibly deploying AI with stronger safeguards in governance, transparency, audit processes, and cybersecurity. India has also strengthened cyber supervision of regulated financial entities and expanded digital-fraud monitoring systems. Yet major gaps remain. Smaller cooperative banks and fintech firms often lack sophisticated cyber defences. Dependence on foreign Cloud infrastructure and frontier AI systems creates strategic vulnerabilities.

 

The challenge is not only national but global. Cyber threats may not respect borders. An attack on a global Cloud provider or payment network can spread rapidly across jurisdictions. Emerging economies are especially vulnerable because they often lack adequate cyber capacity. This is why the IMF has emphasised intelligence sharing and coordinated regulatory standards. Importantly, AI itself can be part of the solution. Financial institutions are increasingly using AI tools for fraud detection, threat monitoring, vulnerability assessment, and incident response. AI can help identify weaknesses before hackers exploit them and improve resilience through faster recovery systems. But such technological solutions alone are insufficient. Cyber resilience requires continuous stress testing, board-level oversight, redundancy in critical systems, and human supervision. Hence, the growing fusion of AI and finance demands global governance frameworks and preparedness to deal with threats.