Global Regulators Monitor Anthropic’s Mythos AI Over Cyberattack Risks

Central banks in Australia and New Zealand are closely monitoring Anthropic PBC’s new AI model, Mythos, due to its potential to enable sophisticated cyberattacks, according to separate statements from the institutions. Japan’s Finance Minister Satsuki Katayama is set to meet with the country’s largest banks and financial institutions as early as this week to address the perceived Mythos threat, people familiar with the matter told Bloomberg.

Anthropic announced Claude Mythos Preview on April 7, describing it as a frontier AI model so powerful that the company chose not to release it publicly, prioritizing security over broad access. According to Anthropic, the model can autonomously identify unknown vulnerabilities in widely used systems, generate working exploits, and execute complex cyber operations with little human oversight. Testing revealed numerous weaknesses, though their severity and real-world exploitability require further validation, as noted in reports from the World Economic Forum.

This development marks a pivotal shift in the AI landscape, where deployment constraints stem from security risks rather than commercial factors. Governments and financial regulators worldwide are responding swiftly, with US officials urging major institutions to test advanced AI like Mythos in controlled settings to assess both offensive dangers and defensive opportunities. Cybersecurity experts, including Igor Protasowicki from Vistula University, highlight how such models detect vulnerabilities faster than humans but could also empower cybercriminals targeting weak security postures.

The financial sector faces immediate implications, as banks grapple with the dual-edged nature of frontier AI: a tool for bolstering defenses yet a potential weapon for disruption. Market reactions have been sharp, with fears over Mythos contributing to volatility in global technology stocks and raising concerns about the stability of the digital economy. Australia’s Reserve Bank (RBA) and New Zealand’s Reserve Bank (RBNZ) emphasized ongoing vigilance in their statements, signaling proactive oversight amid these uncertainties.

Looking ahead, these high-level meetings and monitoring efforts underscore the urgency for coordinated action. Financial institutions must evaluate their cybersecurity frameworks against AI-driven threats, while regulators weigh how to harness Mythos-like models for protection without unleashing broader risks. As the situation evolves rapidly—from theoretical concerns to practical global responses—the focus remains on balancing innovation with safeguards for critical infrastructure.