Singapore sounds alarm on AI cyber threat as Anthropic’s Mythos model rattles banks
SINGAPORE, April 20 — Singapore’s financial watchdog has warned banks to urgently shore up their cyber defences, as fears grow over a powerful new artificial intelligence model deemed too risky for widespread release, Bloomberg reported.
The Monetary Authority of Singapore (MAS) said it is working with the Cyber Security Agency of Singapore (CSA) to strengthen protections at critical infrastructure operators, including lenders, amid concerns that advances in AI could supercharge cyberattacks.
“Financial institutions need to redouble efforts to strengthen their security defences, pro-actively identify and close vulnerabilities, and raise vigilance on cyber hygiene, including timely security patching,” a MAS spokesman said.
The warning comes as scrutiny intensifies over “Mythos”, a new AI model developed by Anthropic, which has been restricted from wider release due to its potential risks.
Authorities said rapid progress in artificial intelligence could dramatically change the cyber threat landscape.
“Advances in artificial intelligence will accelerate the discovery and exploitation of software vulnerabilities in IT systems,” the MAS spokesman added.
While Singapore’s CSA did not name the model directly, it issued an advisory on April 15 cautioning about similar risks, signalling heightened concern across the city-state’s digital infrastructure.
The move reflects a broader global unease, as regulators and financial institutions grapple with the implications of increasingly capable AI tools.
According to Bloomberg, discussions are under way between regulators and banks on how firms are managing the cybersecurity challenges posed by such systems.
Anthropic has held back a broader rollout of Mythos after internal testing showed it could identify long-hidden software flaws — a capability that could be weaponised by malicious actors.
The risks have already triggered high-level attention in the United States.
Treasury Secretary Scott Bessent and Federal Reserve chair Jerome Powell recently convened an urgent meeting with Wall Street bank leaders to assess vulnerabilities and safeguards.
Executives are also acknowledging the dual-edged nature of the technology.
Jamie Dimon, whose bank is testing Mythos, said on a recent earnings call that artificial intelligence has intensified cyber risks, even as it offers new tools to strengthen defences.
