AI cyberattacks: G7 prepares financial sector exercises
· BehindLines newsroom
G7 finance ministers and central bank governors agreed to prepare exercises simulating cyberattacks that use artificial intelligence. Japan's finance minister, Satsuki Katayama, confirmed the decision after a meeting held in Bangkok on the sidelines of the International Monetary Fund's (IMF) annual meetings, according to Reuters and Techzine.
Simulating a crisis without attacking real systems
These will be "tabletop" exercises: officials work through hypothetical scenarios without attempting to penetrate or disrupt a bank's operational systems. This allows them to test alerts, information-sharing, coordination among institutions and the measures required to restore a critical service.
G7 finance ministries and central banks would form the initial group of participants. Private financial institutions could be brought in later, depending on the scenarios chosen. No dates, simulated incidents or definitive list of participating institutions have been announced.
The objective is not to demonstrate that a particular cyberattack has already succeeded. Rather, the findings should help assess existing arrangements and, if necessary, update incident-response manuals and procedures.
A warning from the IMF
The decision follows a warning from the IMF, reported by Reuters that same week: an AI-enabled cyberattack on a critical service at a financial institution could undermine confidence in the financial system. This is a potential risk, not evidence that such an attack has caused an actual banking outage.
AI can, for example, help generate more convincing fraudulent messages or speed up some stages of an attack. Conversely, banks can also use it to detect suspicious activity; the announcement of these exercises alone does not establish how effective such defences are.
Because payment and financing systems are interconnected, the disruption of an essential service can affect organisations beyond the bank or country where it occurs. The G7 simulations are intended to help authorities identify these dependencies before a real crisis takes place.