Opinion: Anthropic’s Mythos is a warning shot. Singapore’s banking system needs to be ready

Frontier AI is no longer just a story about productivity tools or consumer applications. It is becoming a question of critical infrastructure, cyber resilience and financial stability, says NTU Professor Lin William Cong, from the Nanyang Business School.

In a Straits Times commentary, Prof Lin wrote that the emergence of Anthropic’s Claude Mythos Preview model shows why financial centres such as Singapore should pay attention to how advanced AI could affect banking and digital payments.

Anthropic says Mythos has discovered vulnerabilities in major browsers and operating systems, including weaknesses in foundational digital infrastructure. Rather than release the model broadly, the company is reportedly offering it first to major technology and infrastructure firms so they can patch their systems before adversaries acquire similar capabilities.

Prof Lin said reasonable people can debate whether Anthropic is overstating what Mythos can do. But for policymakers, the key issue is not whether every claim about the model is fully proven, but that the possibility is being taken seriously by government officials and major financial institutions.

He added that Singapore should care early, not late. As a major financial hub and a regional base for global banks, Singapore would not be insulated from a serious AI-driven cyber incident affecting international finance.

If more powerful AI tools make it easier to find software weaknesses, automate attacks or exploit common digital systems used by many organisations, the effects will not stop at banks or regulators. They could reach the public through more convincing scams, delayed digital payments or disruptions to banking services.

Read more in The Straits Times.