Terms of Reference
Commissioned by: IPPF Global, Performance, Learning and Impact (PLI), with IPPF European Network (EN)
Budget: Max. EUR 15,000 for core scope (all-inclusive).
Duration: 12 Oct 2026 – 1 Apr 2027. Contract signed week of 12 Oct 2026.
Location: Home-based / remote; all MA engagement online.
Deadline to submit: Friday 28 August 2026, 23:59 CET, to agasser@ippf.org and djagalo@ippf.org, subject “IPPF AI for Research Guidance”
What to submit: One technical proposal (max. 10 pages, excl. CVs/work samples) + one financial proposal in EUR. See Section 5 and 6.
Questions: Written questions by 13 Aug 2026; consolidated answers shared by 17 Aug 2026.
Interviews: Shortlisted candidates interviewed online, week of 28 Sept 2026.
The assignment has three connected components:
1. Federation-wide AI-for-research guidance - practical, principles-first, with a tool annex, worked examples and cheat sheets, grounded in external good practice and piloted in EN (Oct 2026–Apr 2027);
2. Global MA consultation - consultations with a sample of MAs in the other five IPPF regions: Americas and Caribbean; Africa; South Asia; East, South East Asia and Oceania; and Arab World regions (Oct 2026–Jan 2027).
3. EN Member Association AI mapping - anonymized mapping of AI practice/needs across European Network Member Associations (EN MAs), AI-for-research in depth, plus a risk/mitigation matrix (Oct–mid-Dec 2026).
Background
IPPF is a global SRHR service provider and advocate, a movement of national Member Associations (MAs) in 150+ countries, supported by a central Secretariat and six regional offices. IPPF EN is the regional office for Europe and Central Asia.
AI is an IPPF priority for 2026–27. EN has already set the scene (a Use of AI in IPPF EN Publications standard, an AI-for-research tools database, an in-house MEL AI system, and a Federation-wide AI inventory). Global PLI now commissions Federation-wide AI-for-research guidance, piloted in EN, as the template for future MEL-function guidance (2027). The pilot runs in EN due to existing alignment and co-funding for the MA mapping. A wealth of AI-for-research guidance already exists externally (universities, funders, charities, think tanks, publishers) - the consultant adapts it to IPPF’s values, context and governance rather than starting from scratch.
Governance & definitions. The guidance must align with the IPPF IT Policy 2025 (AI usage 2.7, data privacy 2.4, Data Classification 3.4, DPO 3.5), the IPPF Data Management Strategy, the IPPF Privacy Notice, EN’s AI-in-Publications standard and the IPPF Safeguarding Policy and hold across jurisdictions (EU/UK GDPR and comparable regimes – Annex below). “Research” is defined widely: the full research life cycle plus MEL and everyday evidence work. The scope of AI for research will be agreed at inception, with worked examples, starting from the broad working definitions in this ToR (AI, large language models, algorithms; stand-alone and embedded tools).
Objectives
- Define the scope of AI for research at inception, with worked examples and the service-user data boundary.
- Map current AI practice and needs across IPPF EN Member Associations, surfacing shadow use non-punitively – the evidence base for the guidance.
- Consult a sample of MAs in the Americas and Caribbean; Africa; South Asia; East, South East Asia and Oceania; and Arab World regions so the guidance is Federation-wide.
- Develop practical, principles-first, technology-agnostic guidance, co-created and validated with an MA reference group, with an updatable tool annex.
- Ensure data protection and safeguarding hold across jurisdictions (legal layer supplied by in-house Legal/DPO (Data Protection Office).
- Set the guidance up as a living document: named owner, six-monthly review, light adoption pathway.
Scope of Work
- Inception: agree the definitive in/out list of research tasks, tool access (incl. whether a locally run/offline model is needed), the mapping instrument, good-practice source list, and reference-group/validation calendar.
- Service-user-data boundary: aggregated data is in scope; client-level records/interview material are out of scope (Confidential/High-risk under the Data Classification Framework; special-category under GDPR Art. 9). The guidance states the boundary and the route to lift it later (named use case, DPIA, ethics/DPO/MA sign-off, in-house model).
- Mapping A (EN MAs): structured interviews with 6–8 purposively sampled MAs plus EN Secretariat, 2–3 task walkthroughs, and an EN-wide short survey; non-punitive framing; produces a risk/mitigation matrix. Good practice named only with consent; shadow use, incidents and risks never attributed.
- Mapping B (other 5 regions): one-to-one consultations with two MAs per region on current practice and support needs.
- Guidance development: builds on a rapid desk review of external AI-for-research guidance; enabling and principles-first (sanctioned toolset incl. locally run models, prohibited uses/red lines aligned to the Data Classification Framework, human-in-the-loop as default control); plain language, modular, with decision trees, comparison tables and one-page cheat sheets.
- Cross-jurisdiction framework: consultant aligns with EU AI Act, OECD, UNESCO, Council of Europe and WHO standards; IPPF Legal/DPO supply the legal sign-off (GDPR, POPIA, LGPD, etc.).
- Stakeholder validation (2 workshops) and Membership adaptation tooling: MA-adaptable format for later cross-region validation (optional contract extension).
Deliverables & Timeline
- Inception note: Scope, tool-access check, instrument, source list, calendar - by 23 Oct 2026
- EN MA AI mapping: Anonymized practice/needs summary, risk matrix - draft 10 Dec, accepted 18 Dec 2026
- Global MA AI-for-research report: All regions consolidated - draft Membership adaptation, final with guidance 1 Apr 2027
- Draft guidance: Grounded in desk review + mapping evidence, incl. tool annex - by 18 Feb 2027
- Final guidance: Approved, MA-adaptable, plain language + cheat sheets - by 1 Apr 2027
- Adoption pack: One-pager, briefing, roll-out note, owner, review cadence - by 1 Apr 2027
- Success measures & baseline: Light adoption measures for 2027 extension - by 1 Apr 2027
Each deliverable requires written acceptance by the EN technical lead before payment.
Key milestones: EN fieldwork 19 Oct–30 Nov 2026; other-regions consultations to 22 Jan 2027; validation workshop 1 week of 1 Feb 2027; validation workshop 2 week of 1 Mar 2027; consolidated feedback by 18 Mar 2027.
Approach, Management & Budget
Phases: inception → one MA contact round → iterative drafting with technical-lead review → two-validation workshops → light adoption rollout. IPPF values/ethics govern where external practice and IPPF direction diverge (logged in a comment-response log). Evidence blends the consultant’s expertise, desk review, mapping/consultation findings, and the MA reference group (1–2 MAs/region, 6–12 total), which co-creates and validates both products; IP remains with IPPF. The consultant runs all online meetings; IPPF EN provides coordination, introductions and access.
Management: reports to Ane Gasser (EN technical lead, sign-off); Stefanie Wallach (Global PLI) is commissioning lead; Dome Jagalo (Health Research Scientist, PLI) is Global co-lead. Weekly 30-min check-in plus a review point per deliverable. An advisory group (PLI, medical, IT, risk, EN) and the MA reference group accompany the work, and both join the validation workshops.
Budget and Payment: EUR 15,000, all-inclusive, paid in instalments against written deliverable acceptance (one instalment on acceptance of the EN Member Association AI mapping in Dec 2026).
Data protection & ownership: all outputs are IPPF property (IP assigned to IPPF). Consultant handles data per IPPF IT Policy/Privacy Notice/GDPR, signs a data processing agreement, stores materials on EN SharePoint (not personal drives), deletes local copies at contract end, keeps all information confidential, and never enters personal/sensitive IPPF data into non-enterprise AI tools; AI use in producing deliverables is disclosed.
Consultant profile, Submission & Evaluation
Eligibility:
- Relevant degree/experience; track record producing practical guidance/policy for multi-country or federated organisations;
- AI and research-methods expertise able to translate technical material for non-specialists;
- Familiarity with the external AI-in-research guidance landscape;
- Experience designing organizational mappings/needs assessments and facilitating participatory online workshops;
- Familiarity with IT/data-protection governance across regions;
- Safeguarding/research-ethics awareness; strong written English;
- SRHR/international development experience as an asset.
Open to individuals, teams and firms; conflicts of interest must be declared.
Submission:
1. Technical proposal (max. 10 pages excl. CVs/samples – applicant details, methodology, workplan with days per deliverable, assumptions/risks/mitigation, ≥2 work samples + 2 references, CVs max. 2 pp each)
2. EUR financial proposal (core-scope total within the budget cap, cost breakdown by deliverable and days. VAT/GST states either "I am a VAT or GST payer, and my registration number is [number]" or "I am not a VAT or GST payer". Any applicable VAT is shown separately in the table and sits within the maximum available budget to both emails above by 28 Aug 2026, 23:59 CET. Financial offer template provided in the PDF Terms of Reference file.
Evaluation: 100 points total
Step 1 - Compliance check to verify if bidder meets the administrative criteria (received by deadline, technical proposal max 10 pages, excl. CVs/work samples, and financial proposal (in EUR using table above, VAT/GST statement included); key personnel CVs attached (max 2 pages each); minimum two prior work examples and two references provided);
Step 2 - Technical (90 pts, ≥70% required to proceed);
Step 3 - Financial (10 pts, core scope only, lowest compliant price scores full marks, others pro-rata);
Step 4 - Interviews with top 3-4 applicants (week of 28 Sept 2026), which may adjust technical scores.
Contract awarded to the highest combined score. See PDF Terms of Reference for criteria and score card.
Safeguarding & Ethics
Consultant follows the IPPF Safeguarding Policy and Code of Conduct; any mapping and consultations use informed consent, voluntary participation and anonymization of shadow use/incidents/risks; need for formal ethics review is confirmed by the technical lead at inception before data collection begins.
