The Governance Paradox: Rules First, then AI? Or the Other Way Around? How We Have Built the Digital Agent at DKB and Optimised Governance Along the Way.

The rise of AI in banking raises a key paradox: do we need governance in place before we innovate, or can governance adapt as we move? In this keynote, I will share how we at DKB tackled this question while building our Digital Agent. Instead of waiting for a “perfect” rulebook, we applied a hybrid governance model: central standards for compliance and ethics, combined with decentralized flexibility for rapid development. This approach allowed us to learn fast, mitigate risks early, and embed trust into our systems without slowing innovation.

I will show how we balanced regulatory expectations, data protection, and explainability with the need for speed and efficiency. Along the way, we discovered that governance is not a blocker but an enabler if designed pragmatically. The case illustrates how banks can create scalable AI adoption frameworks while staying competitive in a fast-changing market.