Regulators said on Monday they are monitoring the development of Anthropic's frontier AI model Mythos, which experts ​say could have the capability to be used ‌to destabilise banking systems.