Financial regulators worldwide are intensifying efforts to harness artificial intelligence as a defensive weapon against rapidly evolving cyber threats that endanger banking systems and investor protection. Marlene Amstad, president of Switzerland's Financial Market Supervisory Authority (FINMA) and chair of an influential international supervisory technology forum, emphasized in recent remarks that the financial sector faces an urgent imperative to adopt advanced technological solutions to address emerging vulnerabilities before malicious actors can exploit them at scale.
The challenge confronting regulators reflects a fundamental asymmetry in the digital arms race. As criminal hackers and state-sponsored cyber units leverage AI to identify and weaponize software flaws, traditional defence mechanisms have struggled to keep pace with the velocity and sophistication of modern attacks. Amstad underscored that financial institutions must dramatically accelerate their capacity to identify system weaknesses and deploy security patches, compressing what has historically been a slower remediation cycle. The gap between vulnerability discovery and remediation represents a critical window of exposure that modern adversaries increasingly exploit to breach networks and extract sensitive data.
Recognising this imperative, FINMA has taken a leadership role in establishing a coordinated international response. Working through the International Organization of Securities Commissions, a global standard-setting body with regulatory authority spanning roughly 95 percent of worldwide securities markets, the Swiss regulator has championed the creation of a dedicated forum focused specifically on encouraging supervisory agencies to integrate artificial intelligence into their oversight operations. This institutional mechanism brings together the largest financial regulators globally, creating a formal structure for sharing best practices, collaborative development of supervisory tools, and alignment on emerging technological standards.
The practical dimension of this collaborative effort became evident this week when approximately 100 policy specialists and technology experts convened for a hackathon designed to jointly develop new supervisory instruments. Participants focused initial efforts on building artificial intelligence applications tailored to the unique challenges of cryptocurrency and digital asset market supervision, sectors where regulatory innovation has lagged behind rapid asset class growth. By pooling expertise across different jurisdictions and regulatory frameworks, participants sought to accelerate tool development while ensuring that solutions reflect diverse regulatory philosophies and market structures across major economies.
Beyond cryptoasset supervision, regulators are contemplating more fundamental architectural changes to financial infrastructure itself. Amstad indicated that supervisory agencies are actively exploring the possibility of embedding safety mechanisms and compliance safeguards directly into the underlying technical systems that power digital assets and blockchain networks. Rather than implementing oversight primarily through external monitoring and enforcement, this approach would encode regulatory requirements into the systems themselves, making circumvention technically difficult or impossible. Such innovation represents a significant conceptual shift from traditional regulatory frameworks that have relied predominantly on surveillance and sanctions.
The emergence of artificial intelligence as both threat and solution has exposed previously underappreciated operational vulnerabilities within financial institutions themselves. Amstad noted that deployment of sophisticated AI models, including systems developed by companies such as Anthropic, has revealed gaps in institutional readiness to manage AI-related risks. These operational vulnerabilities extend beyond traditional cybersecurity concerns to encompass questions of algorithmic transparency, accountability for AI-driven decisions, and the capacity of human supervisors to understand and oversee increasingly autonomous systems. Financial institutions must develop new governance frameworks and risk management disciplines adapted to the specific characteristics of artificial intelligence systems.
The geopolitical dimension of AI development has further complicated the regulatory landscape. The United States government recently instructed Anthropic to halt international sales of its latest model iterations, citing national security considerations and concerns about potential weaponization of frontier AI capabilities. This export restriction reflects broader tensions between the accelerating pace of AI capability development and government desires to maintain technological advantage and prevent proliferation of the most powerful systems to potential adversaries. For smaller economies like Switzerland and the broader Southeast Asian region, such restrictions create a challenging environment where access to cutting-edge AI tools becomes constrained by superpower geopolitics.
China has responded to such restrictions by accelerating domestic development of alternative systems. 360 Security Technology, a prominent Chinese cybersecurity firm, announced this week the completion of a domestically produced artificial intelligence model positioned as a functional equivalent to leading Western systems. This parallel development trajectory underscores how export controls and geopolitical competition are driving a fragmentation of the global AI ecosystem, with different technological ecosystems emerging across major power blocs. For financial regulators, this fragmentation presents both challenges and opportunities, potentially forcing regional specialisation in AI development while also creating redundancy and reduced single points of failure.
Amstad emphasised that Switzerland and other advanced financial centres must maintain access to the most sophisticated artificial intelligence systems available globally, arguing that supervisory agencies cannot effectively protect financial systems against threats posed by frontier AI capabilities if regulators themselves lack access to equivalent tools. This position reflects growing recognition among policymakers that regulatory effectiveness increasingly depends on technological parity with both financial institutions and threat actors. Without access to state-of-the-art AI systems, supervisors risk falling permanently behind the curve, unable to detect vulnerabilities or anticipate risks that more advanced systems could identify.
The regulatory community increasingly views artificial intelligence not merely as a risk factor requiring management, but as an essential tool for hardening financial systems before they face real-world threats. Amstad highlighted that AI models can be deployed in testing and development environments to systematically search for vulnerabilities, identify potential failure modes, and stress-test systems under conditions that would be impractical or dangerous to simulate using traditional methods. This forward-looking approach to system security reflects a maturation in how regulators think about technology deployment, moving from a posture of restricting and controlling technology toward one of strategically leveraging advanced tools to strengthen systemic resilience.
For Malaysia and other Southeast Asian financial centres, this global regulatory shift carries significant implications. Regional supervisors face pressure to develop comparable AI capabilities to maintain competitive standing in an increasingly technology-driven financial landscape. The cost and complexity of developing sophisticated AI infrastructure may favour larger, wealthier markets, potentially creating a two-tier regulatory environment where well-resourced economies deploy advanced supervisory technology while smaller markets struggle with legacy systems. Regional cooperation and shared investment in AI development infrastructure may offer a viable path for smaller economies to access capabilities that individual countries could not independently develop.
