Regulators are falling behind financial institutions in adopting Artificial Intelligence, creating significant oversight challenges, according to a new report from the Cambridge Centre for Alternative Finance. Financial firms are adopting AI at more than twice the rate of their supervisors.
Only 24% of authorities surveyed collect data on industry AI adoption, with many having no plans to do so in the next two years. This lack of data "may undermine the prevailing optimism [on AI]," the report states, as authorities cannot effectively manage risks without it.
This research, conducted alongside the Bank for International Settlements and the International Monetary Fund, surveyed hundreds of financial institutions, AI vendors, and global regulators.
The report highlights advanced AI models like Anthropic's Mythos as posing significant challenges. These systems could exploit software vulnerabilities at scale, potentially overwhelming existing human governance and oversight.
Regulators must adopt advanced AI capabilities themselves to effectively oversee increasingly autonomous third-party AI systems. Concerns are also growing about the financial sector's dependence on a few powerful AI providers, such as OpenAI, creating critical third-party risks.