Draft
Draft a Do-No-Harm Protocol
Draft a do-no-harm risk identification and mitigation protocol for an evaluation or research activity.
||
This prompt involves designing protections for sensitive populations or sensitive data. Do not include identifying details of real respondents in your AI inputs.
You are an expert in research ethics and do-no-harm protocol design. Draft a do-no-harm risk identification and mitigation protocol for a program or research activity. Tailor risks and mitigations to the population, context, and data collection methods involved.
**Draft Requirements:**
1. **Risk identification breadth.** Identify physical, psychological, social, and economic risks separately for participants, communities, and data collection teams. Do not default to a single risk type.
2. **Context-specific risks.** Name risks particular to the operating context (conflict, marginalization, surveillance, displacement) rather than relying on generic risk language.
3. **Mitigation specificity.** For each identified risk, name a concrete mitigation step so the protocol is operationally usable.
4. **Escalation and response.** Specify procedures for incident response and escalation, including who is contacted, on what timeline, and what immediate action is taken.
5. **Monitoring of risks.** Build risk re-assessment into the activity timeline so emerging or changing risks are caught during implementation, not only at design.
**Output Format:**
Produce:
1. A 1-paragraph context statement summarizing the activity and the highest-stakes risk categories.
2. A risk register table: risk ID, risk type (physical/psychological/social/economic), affected party (participants/community/team), description, likelihood, severity, mitigation step, monitoring trigger, responsible person.
3. An escalation and response protocol with named roles, contact procedures, and timelines for each incident type.
4. A risk monitoring schedule tied to the activity timeline (pre-deployment, weekly during data collection, post-collection debrief).
5. A 'stop-collection' decision rule identifying the conditions under which data collection is suspended.
Review the outputDo No Harm Protocol
draftdo-no-harmethicssafeguardingrisk-mitigation
Scoring Rubric
Do No Harm ProtocolUse this rubric to score and improve the AI output from this prompt.
Related Prompts
Learn More
Related Guides
- How to Govern AI Use in Your M&E Work
Most AI failures in M&E are not run-time problems. They are setup problems. Before AI touches the work, six decisions shape whether the outputs will hold up: which tasks AI should handle, where your data goes, where humans stay in the loop, and when to stop.
- How to Protect Data Privacy When Using AI for M&E
Beneficiary data belongs to beneficiaries, not AI servers. The SAFE Framework helps you use AI tools without risking a data protection breach, donor compliance violation, or loss of community trust.