Responsible AI: Ethics in Software Development
summary
Microsoft’s approach centers on six Responsible AI principles — fairness, reliability and safety, privacy and security, inclusiveness, transparency and accountability — and shows concrete fixes: a loan risk pilot where the model initially approves only male borrowers because training data reflects 200 years of bias, a staged deployment for a mining driver‑alert system, a healthcare risk model trained on 32,000 patient records with explainable score breakdowns, and a Twitter chatbot that learns abusive speech within 24 hours, demonstrating before/after demos and human‑in‑the‑loop audits.
AI-written summary · about
chapters (14)
- 0:00Intro
- 2:54The reality of implementation is very complex
- 3:59The Evolution of Solutions as we know them
- 8:34Diversity Issues
- 10:03Accountability Issues
- 11:46Transform every area of your organization by bringing Al to...
- 14:45What are the qualities of an Al-ready culture?
- 18:45Our guiding principles
- 19:25Fairness
- 22:48Reliability and Safety
- 24:26Privacy and Security
- 26:22Inclusiveness
- 28:45Transparency
- 31:31Building a better future with Al
related talks
30mResponsible AI – Principles & Practice
34mResponsible AI: The Key to Ethical, Safe and Inclusive Software Development
03mDeveloping AI responsibly
16mResponsible AI at Scale
40mSocially responsible AI: Practical tools for robust design
12mNavigating the Future: Principles of Responsible AI
27mWriting the Playbook for Fair & Ethical Artificial Intelligence & Machine Learning
37mRoadmap to Responsible AI