AI Governance in Financial Services: What Every Developer Should Know
summary
Explains that the model is one component and everything wrapped around it — identity permissions logs and clear workflows — decides deployability, not accuracy alone; banks enforce a three‑lines‑of‑defense review, independent validation, and monitoring for behavior drift (examples: agent actions vs summaries, 92% accuracy failing when chained); shows concrete controls to build now — kill switch and owner, audit trails, approval gates — and argues humans must be able to actually stop or change the system.
AI-written summary · about
chapters (12)
- 0:00AI governance intro
- 1:14Why governance matters
- 3:21Governability as requirement
- 6:55Seven pillars of governability
- 8:24Three lines of defense
- 11:43Limitations of human oversight
- 13:09Detecting system drift
- 14:33Rules to principles shift
- 17:46Six auditable AI principles
- 20:35Governance deepens tech understanding
- 22:01Governance as constraint
- 23:05Regulators' future questions
related talks
24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…- 24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…
31mAI risk governance
52mAI Governance: The Security Perspective
08mWhy AI Governance Matters More Than Ever
33mAI in the Security Stack
12mAI Data Governance for Scalable Compliance in Modern Banking
15mAI Security Simplified: Turning Complexity into Cohesion