Regulators Fall Behind as Banks Adopt AI at Twice the Rate
Financial regulators are losing ground in their ability to monitor artificial intelligence risks as banks and fintech firms deploy these systems far faster than supervisors can track them. A study released this week found that only 20% of regulatory authorities report advanced AI adoption, while financial institutions are adopting the technology at more than double that rate.
The research from the Cambridge Centre for Alternative Finance, conducted with the Bank for International Settlements and the International Monetary Fund, surveyed 350 financial institutions, 140 AI vendors, and 130 central banks across 151 countries between October 2025 and January 2026.
Data Collection Gaps Undermine Oversight
Regulators lack basic visibility into how widely AI is spreading through the financial system. Only 24% of supervisory authorities currently collect data on AI adoption within their industries. Another 43% said they have no plans to start collecting this information within the next two years.
"Authorities cannot successfully oversee AI if they are navigating its adoption and risks without hard data," the report said.
This blind spot matters because advanced systems like Anthropic's Mythos model could exploit software vulnerabilities at scale, straining existing governance frameworks that were built for slower-moving technologies.
Accountability Breaks Down With Autonomous Systems
Financial regulators traditionally hold firms accountable for harms, including cyberattacks, regardless of whether AI systems are built in-house or supplied by vendors. That principle becomes harder to enforce when third-party systems operate with significant autonomy.
"The position becomes harder to apply in the context of more autonomous systems that are provided and managed by third-party vendors," the authors wrote.
Concentration Risk in AI Providers
The financial sector's reliance on a handful of AI providers creates systemic risk. Sixty-nine percent of all survey respondents reported using OpenAI, rising to 76% among financial institutions. Just over half used Google models, while slightly more than a third relied on Anthropic systems.
Heavy dependence on a small group of vendors could expose the financial system to disruptions from pricing changes or service outages.
Emerging Markets Face Steeper Challenges
The oversight gap is particularly acute in emerging markets, where regulatory authorities often lack both technical expertise and data infrastructure to integrate AI into their frameworks, according to the World Bank.
The report suggests that regulators themselves may need to adopt more sophisticated AI tools-systems capable of acting independently-to keep pace with the technologies they oversee.
Your membership also unlocks:
