Singapore banks meet MAS over AI risks linked to Anthropic’s Mythos

• MAS meets Singapore bank chiefs over AI cyber risks.
• CSA urges Singapore infrastructure owners to review cyber risks.
• EU and US assess AI model risks.

MAS convened the chief executives of major financial institutions in Singapore to discuss measures against AI-related cyber threats following the limited release of Anthropic’s Mythos model, according to The Business Times.

Senior Minister of State for Digital Development and Information Tan Kiat How said the government is working with partners that have access to Mythos. The aim is to better understand the model’s capabilities. Tan said the government does not have access to Mythos and is not aware of any local bank that has been granted access.

Anthropic has released the model only to a limited set of partners under a controlled preview.

CSA asks boards to review cyber risk

The Cyber Security Agency of Singapore has issued a letter urging important information infrastructure owners to commission a board-level review of their cyber-risk posture.

“The underlying shift is broader and the risks are real. We are treating them with the seriousness they deserve,” Tan said.

Tan described AI-enabled cyber risk as part of existing systemic risk concerns. “We view AI-enabled cyber risk as an amplification of an existing systemic risk, rather than a wholly-new category,” he said.

CSA’s letter was addressed to the boards and senior leadership of important information infrastructure owners. Tan said the issue should not be left to IT teams alone. “It demands leadership attention at the highest levels, including board members and chief executives,” he said.

Banks monitor AI-enabled threats

MAS earlier said advances in AI could speed up the discovery and exploitation of software vulnerabilities in IT systems. The regulator said financial institutions should strengthen security defences and close vulnerabilities quickly. It also told them to maintain cyber hygiene, including timely patching.

MAS said it was coordinating with CSA to support important infrastructure operators.

The Association of Banks in Singapore said on April 27 that it was working with member banks to monitor emerging threats, share intelligence, and coordinate risk-mitigation efforts. ABS director Ong-Ang Ai Boon said local banks had enhanced monitoring and incident response capabilities. She said the measures support faster detection, containment, and remediation of threats.

Singapore’s three local banks have also addressed the issue. DBS Group chief executive Tan Su Shan said Mythos “amplifies the risk,” adding that attackers could use such tools to find weaknesses faster, while banks could also use them to defend faster.

“From both a speed perspective, it’s faster to market, and from a volume perspective, the blast radius is fast,” Tan Su Shan said.

Praveen Raina, OCBC’s head of group operations and technology, said AI-related solutions undergo “rigorous assessment and validation before deployment.”

UOB said its use of AI is governed by existing cybersecurity controls and internal guardrails. A UOB spokesperson said the bank takes “a disciplined and responsible approach to innovation.”

Mythos draws scrutiny

The concern centres on Mythos, which Anthropic has described as its most advanced AI model to date. The company said the model is designed for defensive cybersecurity tasks. Anthropic has limited its release because of concerns about potential misuse.

Reuters reported that Mythos is designed to find flaws in computer code. Cybersecurity experts have warned that such capabilities could be used to intensify attacks on banks’ technology systems. The model has not been made available to European banks, according to Reuters.

EU and US review AI model risks

European Economic Commissioner Valdis Dombrovskis said the European Commission is in contact with Anthropic about Mythos and is assessing its implications for EU policies and legislation.

Dombrovskis said Commission representatives had met Anthropic and received a briefing on the model’s cyber capabilities and risks.

In the United States, Reuters reported, citing The New York Timesthat President Donald Trump is considering government oversight of new AI models. The report said the US government is discussing an executive order to create an AI working group. The group would involve technology executives and government officials and examine potential oversight procedures.

A White House official declined to confirm or deny the report. “Any policy announcement will come directly from the president. Discussion about potential executive orders is speculation,” the official said.

Want to learn more about Cloud Computing from industry leaders? Check out Cyber Security & Cloud Expo taking place in Amsterdam, California, and London. This comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.