Japan’s finance minister confirmed on Thursday that G7 economies are planning “tabletop exercises” to simulate AI-enabled cyberattacks against banks.¹ The exercises are intended to test how financial institutions would respond when artificial intelligence is used to accelerate or sharpen a hacking campaign.

The move follows months in which governments and organisations have become increasingly concerned that fast-moving AI tools could supercharge cybercrime. The concern is not merely theoretical: officials are weighing scenarios shaped by major new AI releases and reports of rogue-agent attacks, developments that have made the prospect of automated and more sophisticated intrusions harder to dismiss.¹

The G7’s approach is preventive rather than reactive. By running a crisis on paper before one unfolds in the real world, finance authorities aim to expose weak points in bank decision-making, communications and defences without putting customers or markets at risk.

That urgency has also been echoed by senior financial voices. The heads of the Bank of England and the International Monetary Fund have stressed the need to protect financial institutions as cyber threats evolve.¹ Their warning aligns with the G7’s premise: banks are not simply another target for AI-assisted hackers, but a potential pressure point for the wider economy.

A separate report on the planned drills similarly framed them as simulated AI attacks on banks, underscoring that the exercise is becoming a coordinated G7 concern rather than an isolated national initiative.² The test now is whether rehearsing the threat can help institutions move faster than it.