The International Monetary Fund (IMF) has warned that artificial intelligence is significantly increasing the threat of cyberattacks to the global financial system, as demonstrated by the capabilities of Anthropic’s Claude Mythos Preview, which allows even unskilled attackers to exploit vulnerabilities across major systems. This situation heightens the risk of localized breaches cascading into broader financial shocks due to the interconnected nature of financial institutions. The IMF emphasizes the need for policymakers to prioritize cybersecurity as a core financial stability issue, advocating for stronger resilience standards and international coordination to mitigate these emerging risks.

Dastan: Dastan is the parent company of Decrypt and operates Myriad, a prediction market platform. It is mentioned because users on Myriad were weighing the odds of Anthropic publicly releasing Claude Mythos by a near-term deadline.
Anthropic: Anthropic is an AI company focused on building frontier models with a strong emphasis on safety and controlled deployment. It is relevant here because the IMF cites its Claude Mythos Preview as an example of how advanced AI can lower the barrier to sophisticated cyberattacks.
Claude Mythos Preview: Claude Mythos Preview is Anthropic’s advanced preview model, described in recent coverage as a controlled-release system with strong cybersecurity capabilities. In this story, it is used as evidence that AI tools can help even non-experts identify and exploit software vulnerabilities across major systems.
International Monetary Fund: The International Monetary Fund is a global institution that monitors macroeconomic and financial stability and advises policymakers on systemic risks. In this news item, it is warning that AI-enabled cyberattacks should be treated as a financial stability issue rather than a narrow technical problem.

AI cyber risk: Recent IMF commentary frames AI-assisted hacking as a systemic financial stability threat because a single breach can spread across interconnected institutions.
AI defense trend: Financial institutions are increasingly pairing AI-driven attacks with AI-assisted detection, fraud prevention, and incident response tools.
Controlled release: Anthropic has handled Claude Mythos Preview as a restricted rollout, reflecting concerns about advanced offensive cybersecurity capabilities.