Global Watchdogs Struggle to Track Risks From Advanced AI in Finance

A report released Tuesday by the Cambridge Centre for Alternative Finance found that financial firms are adopting AI technologies at more than twice the rate of regulators. Only two in 10 supervisory authorities reported what the study described as “advanced AI adoption,” underscoring concerns about oversight capabilities as increasingly powerful systems enter the market, according to Reuters.

The findings raise fresh questions about whether central banks and financial watchdogs can effectively track emerging threats tied to next-generation AI tools, including models such as Anthropic’s Mythos, which cybersecurity experts have warned could challenge legacy banking infrastructure.

The research, conducted in collaboration with organizations including the Bank for International Settlements and the International Monetary Fund, surveyed a broad cross-section of the industry: 350 traditional financial institutions and fintech firms, more than 140 AI vendors, and 130 central banks and regulators across 151 countries, according to Reuters.

A key issue identified in the report is a lack of reliable data among regulators. Only 24% of authorities said they currently collect information on AI adoption within the industry, while 43% reported having no plans to begin doing so within the next two years. “This empirical blind spot may undermine the prevailing optimism [on AI]. Authorities cannot successfully harness or oversee AI if they are navigating its adoption and risks without hard data,” the report said.

Regulators and global standard-setting bodies have already increased warnings about the risks tied to AI deployment in financial services. Earlier this month, Anthropic introduced its Mythos model, which experts believe could exploit software vulnerabilities at scale. Such capabilities may strain existing governance frameworks and outpace traditional oversight mechanisms, per a Reuters report.

The study suggests that these advanced systems could soon operate with a level of autonomy that complicates accountability. “Regulators generally maintain the principle that financial firms should remain accountable for harms, including cyberattacks, whether AI is built in-house or supplied by third parties, but that position becomes harder to apply in the context of more autonomous systems that are provided and managed by third-party vendors,” the authors wrote.

To keep up, the report argues that regulators themselves may need to adopt more sophisticated, agentic AI tools—systems capable of acting independently without direct human input—mirroring the technologies they are tasked with supervising.

Harish Natarajan of the World Bank noted that the challenge is particularly acute in emerging markets, where authorities often lack both the technical expertise and the data infrastructure needed to integrate AI effectively into regulatory frameworks, according to Reuters.

Beyond oversight gaps, the report also highlighted growing concerns about concentration risk within the AI ecosystem. A significant majority—69% of all respondents—reported relying on OpenAI, with usage rising to 76% among financial institutions. This heavy dependence on a small group of providers could expose the financial system to disruptions, including pricing volatility and service outages.

At the time of the survey, conducted between October 2025 and January 2026, just over half of respondents reported using models from Google, while slightly more than a third relied on systems from Anthropic.

Related: Reuters