ASIC Joins Global Regulators Monitoring Anthropic’s Mythos AI for Banking System Risks
April 20, 2026 – 9:26 am
Australia’s markets regulator, the Australian Securities and Investments Commission (ASIC) has publicly confirmed its monitoring of Anthropic’s Mythos model development alongside international peers, marking a growing global regulatory response. This follows initial warnings from institutions like the Bank of England, US Federal Reserve, and Treasury Department.
ECB President Lagarde has noted the absence of a governance framework for managing these risks.
ASIC is closely monitoring these developments along with peer regulators to assess possible implications for the Australian market. ASIC engages closely with other regulators, government agencies, and the financial sector to understand and respond to changing technologies.
The regulator emphasizes the need for financial services licensees to remain proactive in protecting their customers and clients.
This statement comes as part of a series of global responses to Mythos, an advanced AI model Anthropic launched on April 7, 2026, under a restricted access program called Project Glasswing. Anthropic claims the model successfully identifies and exploits zero-day vulnerabilities in major operating systems and web browsers, a capability raising concerns among regulators about potential systemic risks if accessed by malicious actors.
Global responses have been swift and coordinated:
- Bank of England Governor Andrew Bailey: Warned that Mythos could “crack the whole cyber risk world open” and called for urgent assessment of its impact on financial infrastructure.
- European Central Bank President Christine Lagarde: Admitted the lack of a governance framework to oversee these risks.
- US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell: Convened an urgent meeting of major bank CEOs to discuss Mythos‘s cyber risk implications.
These regulatory actions reflect growing awareness of the potential challenges posed by increasingly powerful artificial intelligence technologies.