The G7 countries are preparing cybersecurity exercises to assess how financial institutions and regulators respond to attacks involving AI.
According to Reuters, the initiative reflects growing concern that artificial intelligence could make it harder to contain cyber incidents. Such attacks could therefore threaten financial stability.
The G7 finance ministers and central bank governors agreed on the plan during a meeting in Bangkok, held in parallel with the annual meeting of the International Monetary Fund.
According to Reuters, Japanese Finance Minister Satsuki Katayama confirmed that the G7 countries had decided to conduct tabletop exercises. These exercises will focus on the risks associated with advanced AI systems.
Tests do not disrupt operational systems
Unlike conventional penetration tests, tabletop exercises simulate security incidents without disrupting operational systems. In these exercises, participants work through hypothetical scenarios and examine how organizations would identify threats. This helps illustrate how organizations coordinate responses to threats and restore critical services.
The G7 plans to use these simulations to evaluate existing incident response procedures. The findings may lead to revised cybersecurity protocols.
Although ministries of finance and central banks will initially participate in the exercises, commercial financial institutions may also join. This depends on the scenarios being tested.
AI brings new risks
The initiative follows warnings from the IMF about the consequences of AI-driven cyberattacks on financial infrastructure. Earlier this week, the organization cited attacks on critical banking services as a potential source of disruption to financial markets.
AI poses additional challenges for cybersecurity teams. Attackers may be able to automate parts of their operations, develop more convincing phishing campaigns, and adapt their techniques more quickly. At the same time, increasingly autonomous AI systems create new operational risks within financial institutions.
Regulators are therefore no longer focusing exclusively on conventional cybersecurity threats. For example, they worry that poorly controlled AI agents could cause incidents whose consequences extend beyond individual organizations.
Bank of England explores recovery options
The G7 initiative coincides with national regulators’ efforts to strengthen the financial sector’s ability to recover from major cyber incidents.
In a speech on Thursday, a deputy governor of the Bank of England warned that financial institutions can no longer assume their existing defenses can withstand increasingly sophisticated, AI-driven attacks.
The G7 has not yet announced when the exercises will take place or which specific attack scenarios will be included.