MAS convenes banks as autonomous AI cyber risks escalate
Officials are monitoring Anthropic’s Mythos and its ability to identify vulnerabilities.
The Monetary Authority of Singapore (MAS) has convened the chief executives of major financial institutions to discuss artificial intelligence (AI)-enabled cyber threats.
The move comes as the government warned that advanced models are making cyberattacks faster and more sophisticated, Senior Minister of State for Digital Development and Information Tan Kiat How said.
“We view AI-enabled cyber risk as an amplification of an existing systemic risk,“ Tan told Parliament on 5 May. “The fundamentals to strengthen an organisation’s cybersecurity matter more than ever.“
The remarks came in response to parliamentary questions on advanced models such as Anthropic’s Mythos, which has been reported to identify and exploit software vulnerabilities autonomously.
Tan said the government does not have access to Mythos, and Anthropic has released it only to a limited set of partners under a controlled preview. It is also not aware of any local bank that has access to the model.
He added that Singapore is working with partners with access to the model to better understand its capabilities and implications, whilst maintaining ties with major AI labs and cybersecurity firms to monitor developments in frontier systems.
The minister noted AI is reducing the time needed to identify vulnerabilities, allowing attackers to exploit systems faster than traditional patching cycles can respond.
“What we have not yet seen is fully autonomous AI agents running end-to-end campaigns. But this is a matter of time, given the trajectory of technological developments,“ he added.
Meanwhile, the Cyber Security Agency of Singapore (CSA) will issue a letter to the boards and senior leadership of all Critical Information Infrastructure (CII) owners, outlining expectations for reviewing cyber risk posture.
The CSA is also reviewing cybersecurity standards and obligations for critical infrastructure operators to account for faster attack timelines.
Tan said the government is also building in-house AI capabilities for cyber defence. “These are being piloted within the Government and will be extended to more agencies and CII owners when ready.“