Public lesson preview

Responsible AI for NGOs and local organizations.

Explore the seven-part professional pathway before requesting enrollment. The preview shows the decisions, applied outputs, and evidence standard; private lesson progress, assessments, and certificate controls remain inside the learner portal.

Request professional enrollment Open starter workbook

An NGO team sorting AI use cases into decision zones — **Triage the use**Purpose, affected people, data, and decision impact come first.

An NGO team assigning owners and controls to AI risks — **Own the risk**Controls need owners, evidence tests, and review dates.

Practitioners reviewing an AI-assisted report against sources — **Retain authority**Reviewers must be able to challenge, reject, and stop.

Orientation

From AI enthusiasm to an operating decision

Use a realistic 35-person local NGO case to build a governed adoption pathway.

Define the role of law, policy, ethics, and evidence.
Separate learning from legal or compliance claims.
Set the final readiness-pack standard.

Lesson 1

Map use cases and decision impact

Record purpose, affected people, data, output, failure consequences, and review authority.

Distinguish assistance from consequential decisions.
Use allow, restrict, pilot, or reject decisions.
Document why AI is necessary.

Lesson 2

Set data boundaries before tool selection

Classify personal, sensitive, partner, donor, research, and security information.

Define prohibited and approval-required data.
Test contextual re-identification risk.
Choose safer public, synthetic, or aggregated inputs.

Lesson 3

Build a living AI risk register

Turn broad concerns into failure pathways, affected groups, residual risk, and testable controls.

Assign owners and evidence tests.
Set escalation and stop thresholds.
Review risk after controls.

Lesson 4

Design meaningful human oversight

Give competent reviewers the evidence, time, independence, and authority to intervene.

Match review level to consequence.
Counter automation bias.
Keep final accountability explicit.

Lesson 5

Write staff guidance that survives pressure

Translate governance into approved tools, clear red lines, records, disclosure, and support routes.

Use role-specific examples.
Define incident and exception routes.
Test guidance with everyday questions.

Lesson 6

Build and test the 90-day readiness pack

Assemble the governance, due diligence, pilot, monitoring, and incident evidence leadership needs.

Run three tabletop incidents.
Measure benefit and harm.
Make a recorded scale, restrict, pause, or reject decision.

Teaching boundary

The course teaches governance practice, not blind tool use.

This pathway draws on UNESCO, the EU AI Act, UNHCR, and ICRC guidance. It does not replace legal advice, a data-protection impact assessment, safeguarding or protection review, security testing, donor approval, or formal organizational assurance.

Resources

Start building the evidence pack.

Starter workbookPrintable use-case map, risk starter, and human review checklist. Risk register CSVEditable risk-control template for NGO teams. AI policy outlineEditable structure for an internal AI guidance note. Use-case decision canvasRecord affected people, data, consequences, review, and the governance decision. Human oversight matrixDefine competence, source evidence, intervention authority, and stop conditions. 90-day pilot planPrepare, test, learn, and decide with benefit, harm, and stop measures.