AI Foundations for NGOs — Lesson 2: Data Readiness, Ethics, and Responsible AI

Learning Objectives

• Explain the role of data quality, consent, privacy, and security in the course scenario.
• Apply bias, inclusion, transparency, and accountability through a guided implementation or analysis.
• Validate human review and escalation using documented evidence and acceptance criteria.
• Troubleshoot risks, quality issues, and operational gaps related to vendor and tool evaluation.

Detailed Topics

• Data Quality, Consent, Privacy, And Security: Apply the concept to requirements, configuration, analysis, controls, and expected outcomes.
• Bias, Inclusion, Transparency, And Accountability: Apply the concept to requirements, configuration, analysis, controls, and expected outcomes.
• Human Review And Escalation: Apply the concept to requirements, configuration, analysis, controls, and expected outcomes.
• Vendor And Tool Evaluation: Apply the concept to requirements, configuration, analysis, controls, and expected outcomes.

Guided Demonstration

1. Review the scenario, users, data, dependencies, permissions, risks, and expected result.
2. Demonstrate the relevant configuration, analysis, design, administration, or development workflow step by step.
3. Capture settings, outputs, logs, screenshots, calculations, tests, and other useful evidence.
4. Compare the observed result with acceptance criteria and explain corrective actions.

Hands-on Exercise

1. Translate the scenario into a short implementation or analysis plan.
2. Complete the workflow using the supplied checklist and secure working practices.
3. Test normal, edge, and failure conditions and record the results.
4. Document one improvement that increases quality, security, reliability, usability, or business value.

Validation Checklist

• Requirements, assumptions, owners, and success criteria are documented.
• Configuration or analysis results are supported by reproducible evidence.
• Security, privacy, accessibility, quality, and governance controls are reviewed.
• Errors and unexpected results are investigated before approval.
• The final outcome and next action are communicated clearly.

Troubleshooting and Improvement

Use logs, formulas, queries, settings, permissions, test outputs, and stakeholder feedback to isolate the cause of a problem. Correct the smallest verified cause first, retest the complete workflow, document the evidence, and add the learning to the final improvement plan.