What is AI Ethics & Governance in Projects?
When projects use or deliver AI, governance questions become project management questions: data privacy (what may feed the model?), bias (who does the system fail?), transparency (can decisions be explained?), accountability (a human owns every consequential output), and compliance with fast-moving regulation.
Exam posture: the PM doesn't need to build models, but must ensure ethical review is in the plan — data governance in requirements, bias testing in quality, AI risks in the register, and no confidential project data pasted into unapproved tools.
Worked example
A hiring-platform project adds an AI résumé screener. The PM treats ethics as scope: bias testing across demographic slices goes into the quality plan (finding: the model penalizes employment gaps — disproportionately affecting caregivers; it's corrected), explainability becomes an acceptance criterion, and legal's AI-regulation checklist gates the release. Six months post-launch, a regulator asks how decisions are made — and the project has an answer instead of a scandal.