The International Monetary Fund (IMF) warned Thursday that artificial intelligence is rapidly amplifying cyber threats to the global financial system. In a new blog post, IMF economists highlighted Anthropic's Claude Mythos Preview AI model, which was found capable of identifying and exploiting vulnerabilities across all major operating systems and web browsers, even when used by non-experts. This demonstrates how AI is lowering the barrier to sophisticated cyberattacks. The IMF cautioned that AI may concentrate risk across the financial system, as heavy reliance on a few cloud providers, software platforms, and AI models means a single exploited weakness could cascade across many institutions simultaneously. Such scenarios could elevate cyber incidents from operational issues to macro-financial shocks, triggering confidence crises, liquidity strains, and market fire sales. The geopolitical dimension adds complexity, as inconsistent oversight across countries weakens the interconnected financial system, with emerging economies facing disproportionate exposure. For wallet and key holders, this underscores the urgency of using secure, audited wallets and following best practices like hardware wallets and multi-signature setups. Users should remain vigilant against AI-powered phishing and social engineering attacks, which are becoming more sophisticated. The IMF urges policymakers to treat cybersecurity as a core financial stability concern, but individual users must also take proactive steps to protect their assets in an environment where AI-powered threats are evolving rapidly.

Key facts

• IMF warns AI lowers barriers to sophisticated cyberattacks on financial systems.
• Claude Mythos model found to exploit vulnerabilities across all major OS and browsers.
• Single breach could cascade across interconnected financial institutions.
• AI is both a threat and a tool for defense in financial cybersecurity.
• Inconsistent global oversight worsens cyber risks for emerging economies.