The Monetary Authority of Singapore (MAS) is urging banks to bolster their cybersecurity measures in response to rising concerns over the deployment of Anthropic’s Mythos AI model throughout Asia. This follows warnings from US officials and international regulators regarding the potential for AI-driven cyberattacks, which are exacerbated by advanced models like Mythos, known for its superior capabilities in coding and cybersecurity research. In light of these developments, the MAS and the Cyber Security Agency have recommended that financial institutions improve their cyber hygiene practices.

Anthropic: Anthropic is an AI safety-focused company developing advanced large language models under the Claude family, emphasizing reliable and interpretable AI systems. Its latest Mythos model, also called Claude Mythos Preview, excels at autonomously identifying and chaining software vulnerabilities in major operating systems and browsers. Anthropic has limited Mythos access to select partners via Project Glasswing to enable defensive cybersecurity improvements while mitigating offensive misuse risks.
Monetary Authority of Singapore: The Monetary Authority of Singapore (MAS) functions as the central bank and primary financial regulator, overseeing monetary policy, financial stability, and supervision of banking institutions. Recently, MAS issued guidance urging banks to improve cybersecurity practices, including proactive vulnerability scanning and timely patching. This response addresses emerging risks from advanced AI models capable of exploiting software weaknesses.

`json
{
“AI Capabilities”: “Mythos is an AI model developed by Anthropic, known for its advanced hardware utilization and capabilities in research related to cybersecurity.”,
“Regulatory Alert”: “Singapore’s Monetary Authority has recommended financial institutions to strengthen their cybersecurity measures in response to potential threats related to Anthropic’s Mythos model.”
}
`