A retail bank relies on AI models across a growing number of business use cases. When one model is found to have been compromised, leaders must rapidly determine the extent of the impact, understand which business processes may be affected, and decide how best to respond.

This presentation follows the bank's journey as it works to strengthen its approach to AI governance in response to this challenge. While established governance and model risk management processes are already in place, the organisation faces increasing complexity as the number of models, use cases and stakeholders continues to grow. The bank recognises the need for a more connected understanding of its AI landscape, including model dependencies, business impacts and ownership, to support transparency, accountability and regulatory readiness.

One Poisoned Model. Two Affected Use Cases. One Urgent Decision.

Speakers

Megan Burns