Australian Regulator Urges Financial Sector to Address Cybersecurity Risks Posed by AI Models Like Mythos

Australia's corporate regulator has called on the nation's financial industry to take urgent action to address the cybersecurity risks potentially introduced by advanced artificial intelligence systems such as Mythos.

The Australian Securities and Investments Commission (ASIC) released a letter to the financial services sector on Friday, emphasizing that additional measures are necessary to ensure cybersecurity protocols are as robust as possible.

"Cybersecurity risks have entered a new era. The emergence of cutting-edge AI models presents both opportunities and significantly heightened risks, exposing vulnerabilities faster than many anticipated," stated ASIC Commissioner Simone Constant.

"Do not wait for complete clarity on the threats posed by new AI models. Instead, act now and proceed rigorously to strengthen the cyber resilience foundations supporting your business," she added.

Experts have warned that Mythos, with its advanced coding capabilities, could potentially identify cybersecurity vulnerabilities at an unprecedented scale, representing a substantial risk.

Anthropic, the developer of Mythos, has not yet responded to ASIC's letter.

ASIC's warning follows a statement last month from Australia's banking regulator, which indicated that information security measures within the domestic financial services industry are struggling to keep pace with the rapid development of artificial intelligence.

"Time is of the essence—if your cybersecurity defenses are not already in place, now is the moment to act and prepare," Commissioner Constant emphasized.

Anthropic has released a preview of Claude Mythos under "Project Glasswing," a program with highly restricted access that includes major technology firms such as Amazon, Microsoft, Nvidia, and Apple.

A recent survey revealed that regulators are significantly lagging behind financial institutions in the adoption and understanding of artificial intelligence. A lack of data on emerging threats has also raised concerns about the ability of central banks and financial watchdogs to effectively monitor and respond to risks associated with AI models.