3.1. Responsible AI Principles
💡 First Principle: Responsible AI is about managing a tool that is powerful but fallible — so the principles all reduce to one stance: keep a human accountable for outcomes the AI only suggests. Fairness, safety, privacy, transparency, and accountability are the dimensions of that stance.
Why this matters: the exam frames responsible AI around recognized principles (the kind Microsoft and GitHub publish), and asks you to apply them to coding situations. You need to recognize a risk when a scenario describes one, name an appropriate mitigation, and reject answers that abdicate human responsibility.
The mental model: think of Copilot like a powerful power tool. The tool isn't "ethical" or "unethical" on its own; responsibility lives in how it's operated — guards in place, operator trained, output inspected. Responsible AI principles are the safety standards around the tool.
⚠️ Common Misconception: "Responsible AI is just about avoiding offensive output." Responsible use is far broader: validating correctness, mitigating bias and harm, protecting privacy and intellectual property, maintaining human oversight, and operating within policy. Tone is the smallest part.
