Humanitarian AI Research Fellowship

Organization: NetHope Closing date: 9 Oct 2026 About the Fellowship NetHope is piloting a new, short-term Humanitarian AI Research Fellowship to grow the pool of researchers and practitioners who understand how AI can responsibly advance humanitarian and development work, and to produce research findings that NetHope's Member organizations and the wider nonprofit community can put to immediate, practical use. This is a part-time, remote, internationally open opportunity for individuals who bring rigor, curiosity, and a genuine interest in responsible AI for humanitarian impact. About NetHope NetHope brings together leading humanitarian, development, and conservation organizations to create a more connected and equitable world through the power of technology. At the intersection of the nonprofit, private, and technology sectors, NetHope unites more than 60 Member organizations working in over 190 countries. In collaboration with over 70 of the world’s top technology companies, we design innovative, scalable solutions to global challenges, helping our Members and the wider nonprofit sector better serve people and communities in need. Two Research Streams The fellowship will run two parallel streams. Fellows will work independently within a stream, with guidance and support from the AI Fellowship Lead and periodic check-ins with NetHope subject-matter experts. Stream 1: AI Adoption & Governance Focused on at least two case studies of how humanitarian and nonprofit organizations are adopting, governing, and managing risk around AI tools in their operations and programs. Illustrative project areas include: • Documenting real-world lessons and practices, helpful or cautionary from organizations integrating AI tools into their operations • Building on NetHope's AI readiness work to better capture what adoption looks like in practice at the level of individual implementing partners • Comparing organizational strategies for AI adoption (direct vendor partnerships, open-source and in-house builds, or off-the-shelf tools) and factors that may drive the choice • Identifying priority AI safety and governance risks across humanitarian organizations, and practices for risk identification, mitigation, and incident response Stream 2: AI Systems & Localization Focused on the technical side of AI for humanitarian use, particularly where localized, offline, or low-connectivity deployment matters. Illustrative project areas include: • Evaluating open-source and open-weight models, including localized or fine-tuned versions, for humanitarian use cases • Assessing the relative risks and advantages of local/open systems versus closed, proprietary systems for humanitarian contexts • Surfacing practical use cases and readiness gaps for AI deployment in resource-constrained or low-connectivity settings; developing a small library of simple, user-friendly tools to solve specific problems • Contributing to a library of supervised fine-tuning prompts and datasets for humanitarian contexts Research Standards • Case studies involving NetHope Members or partners require their consent, and NetHope will review all findings before publication. • Fellows will follow NetHope's AI Policy, including completing an AI Disclosure Form before starting, and NetHope's Data Protection Policy. • Research involving personal data or affected communities must meet NetHope's data protection and research ethics requirements. Who Should Apply We welcome applicants from mid-career to seasoned practitioners who bring strong writing and analytical skills and a genuine interest in responsible AI's role in humanitarian impact. • Stream 1 (Adoption & Governance): background in policy, program evaluation, or humanitarian/international development systems; experience with technology adoption or organizational change is a plus • Stream 2 (AI Systems & Localization): background in computer science, machine learning, or applied data science, with knowledge of or interest in the humanitarian sector. Experience in model evaluation and/or red teaming a plus. • Across both streams: strong written communication, self-direction, and the ability to sustain an independent research project on a fixed timeline Fellowship Details • Format: Remote, internationally open, part-time (approximately 10 hours/week) • Duration: Mid-November 2026 through February 2027 • Cohort size: 4-6 fellows • NetHope owns all research outputs and related intellectual property, including any datasets, prompts, tools, and code. Fellows are credited as named authors. Release and licensing terms will be set out in the fellowship agreement. • Fellows receive a stipend in recognition of their time and contributions. Details will be shared with selected fellows and set out in the fellowship agreement. • All fellows must pass an OFAC screening before starting. How to apply f this opportunity interests you, please submit your proposal HERE by October 9th, 2026 Your application should include: • CV • Letter of interest responding to the following questions: • Tell us about your motivations, relevant experience, and interest in working in the humanitarian and nonprofit sector. • In your view, what is one of the top concerns or risks associated with AI in the humanitarian sector, and why? Briefly describe one possible way to mitigate it. • Think of an AI tool (open-source, commercial, or in-house) you've used or evaluated in a real project, not a demo. What was one way it under-performed for the actual context it was used in, and how did you find out? Please don't include confidential information about your employer or clients. • Two to three professional references • Optional: Writing sample Due to the anticipated volume of submissions, NetHope reserves the right to review applications and select fellows prior to the closing date. Only those selected for further consideration will be contacted. NetHope is an Equal Opportunity Employer.