What happened
On 15-16 October 2026, finance ministers and central bank governors of the G7 — meeting on the sidelines of the IMF annual gathering in Bangkok — signed off on running tabletop exercises on frontier AI. Japanese Finance Minister Satsuki Katayama confirmed the drill plan, which will feature simulations developed by G7 finance authorities to test responses to AI-driven cyber attacks and could involve private financial institutions depending on the simulation. Results will be used to test the effectiveness of existing protocols and to revise response manuals. The agreement follows an IMF warning that an AI-enabled cyberattack on a critical financial service could undermine confidence.
Why it matters
This is a concrete step toward coordinated G7 supervisory response to AI-driven cyber risk in the financial sector, and it signals growing regulatory expectation that both authorities and financial institutions rehearse AI-attack response and deep cyber-recovery, alongside the Bank of England's bare-metal recovery work and the ECB's end-October deadline for banking AI-cyber recovery plans.
Action needed
Financial institutions and market infrastructure operators should prepare for heightened supervisory focus on AI-attack resilience and recovery, and be ready to participate in any national/G7 simulations.