AI Guardrails Enable Innovation in Financial Supervision

In this insightful conversation, Oner Can Koltukcu, a policymaker at De Nederlandsche Bank, delves into the evolving challenges of supervising AI within the financial sector. He outlines how financial institutions are rapidly adopting AI technologies—from generative AI in customer service to complex creditworthiness models.

However, the lack of explainability in these systems poses a critical obstacle to compliance, especially when institutions must justify decisions like loan rejections to consumers. This tension illustrates a broader regulatory dilemma: enabling innovation while ensuring accountability and fairness.

Koltukcu emphasizes that robust supervision is not a barrier but a catalyst for innovation. Public trust in AI-enhanced financial services significantly increases when oversight is transparent and well-communicated.

Through initiatives such as cross-functional supervisory teams, sector-wide AI literacy training, and collaborative European guidance, De Nederlandsche Bank aims to build a regulatory environment that supports safe experimentation. The focus is shifting toward asking the right questions—around bias, ethics, and human oversight—rather than prescribing rigid technical standards, thus creating an ecosystem where innovation and regulation co-evolve.