Embedding Non-disruptive Data and AI Governance into Agile Delivery | Aminat Agboola-Osho
As organisations accelerate data, analytics, and artificial intelligence delivery, governance is often experienced as a late-stage approval layer that slows teams down. This session challenges that pattern and presents a practical approach for embedding data and AI governance directly into agile delivery.
The session shows how ownership, data quality, privacy, security, lineage, transparency, and responsible AI controls can be built into discovery, backlog refinement, sprint planning, design decisions, release readiness, and continuous improvement. The aim is to make governance proportionate, decision-led and easy to adopt.
Attendees will learn how to distinguish what AI can do from what an organisation should do, how to keep human accountability visible, and how to measure governance by business outcomes rather than governance activity. The result is trusted innovation: teams move quickly because standards, decision rights and evidence are clear from the start.
What the session covers
Why governance becomes disruptive when it is added after delivery decisions have already been made.
How to embed governance at the point of decision through agile ceremonies and product workflows.
A four-pillar model covering ownership, quality, control, and delivery.
How to separate technical capability from ethical judgement through the Can, Should, Control lens.
How to measure value through customer impact, delivery pace, control confidence and adoption.
Audience takeaways
A repeatable model for making data and AI governance practical, proportionate and delivery-aligned.
Practical prompts for identifying fairness, transparency, accountability and data-quality risks early.
A 90-day starting approach for proving value before scaling governance more widely.
Clear language for positioning governance as an accelerator rather than a bureaucratic layer.