Agile methodology and artificial intelligence are currently on a collision course. "Ship early and iterate" is an important part of Agile, but "safety, fairness, and veracity" is an important part of AI. This creates a problem called "Ethical Debt." Like technical debt, ethical debt builds up without anyone noticing, and it can lead to reputational bankruptcy or regulatory fines under laws like the EU AI Act (up to 7% of global turnover).
This session gives you a practical, expert-level plan for bringing these two sides together. We will break down the idea of Ethical Debt by looking at real-life examples from Amazon and iTutorGroup to show how negligence can cost you money. Next, we'll talk about solutions and show how to "refactor" Agile ceremonies and artifacts for the age of AI. People who come will learn how to:
Put "Fairness Acceptance Criteria" in user stories so that ethics can be tested.

During sprint planning, hold "Consequence Scanning" workshops to find risks early.
Use open-source tools like Fairlearn and MLflow to automate fairness checks and build a "compliance-as-code" pipeline for "CI/CD for Ethics."
At the end of the talk, there is a detailed "Definition of Done" for AI that makes sure that speed doesn't hurt society. Come with us to learn how to go fast without breaking things.