--- title: "Guidelines for Whom? Rethinking — paper by Yoo et al. (2026)" description: "This position paper asks whether AI ethics guidelines for humanitarian work protect refugees when the organizations expected… 2 references." authors: ["Nari Yoo", "Ashley Khor", "Namrata Mukhija", "Aminat Adebiyi", "Miri Zilka"] year: 2026 venue: "Proceedings of the Workshop on Evaluating Evaluations (EvalEval)" source: "https://aclanthology.org/2026.evaleval-1.3/" doi: "10.18653/v1/2026.evaleval-1.3" licence: "CC BY 4.0 (ACL Anthology)" licence_url: "https://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/yoo-2026-guidelines-whom-rethinking-ai-ethics/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/yoo-2026-guidelines-whom-rethinking-ai-ethics/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/35/) / Guidelines for Whom? Rethinking AI Ethics in Resource-Constrained Migration Services # Guidelines for Whom? Rethinking AI Ethics in Resource-Constrained Migration Services **Authors:** Nari Yoo, Ashley Khor, Namrata Mukhija, Aminat Adebiyi and Miri Zilka · **Year:** 2026 · **Venue:** [Proceedings of the Workshop on Evaluating Evaluations (EvalEval)](https://md.research.elamur.ai/free/en/venues/evaleval/) ([EvalEval 2026](https://md.research.elamur.ai/free/en/venues/evaleval/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2026.evaleval-1.3/) · [PDF](https://aclanthology.org/2026.evaleval-1.3.pdf) · [DOI 10.18653/v1/2026.evaleval-1.3](https://doi.org/10.18653/v1/2026.evaleval-1.3) · [OpenAlex](https://openalex.org/W7166841946) **Citations:** 0 citations (OpenAlex, 2026-10-14) ## Summary This position paper asks whether AI ethics guidelines for humanitarian work protect refugees when the organizations expected to follow them are small, underfunded nonprofits. A community group that wants to use AI translation faces several overlapping governance layers at once: the EU AI Act, the IASC Operational Guidance on Data Responsibility, and the Principles for Digital Development. Each was designed with well-resourced governments or large tech firms in mind. The authors draw on participatory fieldwork with refugee-serving organizations in South Korea and two cross-national nonprofit surveys. They argue that when formal compliance is out of reach, AI use does not stop. It moves into informal channels with no oversight, and the cost of having no language access at all goes uncounted. Using realist evaluation (Pawson and Tilley, 1997), they propose judging guidelines by which deployers they work for and under what conditions. They call for requirements that scale with use-case risk and deployer capacity, and for practical deployment-level guidance. Strict oversight would remain for asylum adjudication, biometrics and border surveillance. The argument matters for anyone deploying multilingual or low-resource language tools for vulnerable groups. ## Key points - The number of AI ethics documents grew from 84 (Jobin et al., 2019) to 200 (Corrêa et al., 2023). Most originate in North America and Europe and assume deployers with compliance staff, legal counsel and audit budgets, not nonprofits with three to five employees. - Funding has contracted sharply: - USAID closed in July 2025, with about 83% of its programs canceled and roughly $2.3 billion in migration-related funding lost. - Official development assistance fell 9% in 2024. - As of mid-2025, low- and middle-income countries were home to 71% of refugees. - The paper challenges three assumptions in current guidelines: - that requirements apply equally to all deployers; - that not adopting AI is a neutral, safe default; - that "AI in humanitarian contexts" is one risk category. - The EU AI Act already shows that graduated rules are feasible. It lists migration and asylum AI as high-risk (Annex III, Section 7), while Article 62 gives SMEs simplified documentation and priority sandbox access. - The authors propose three evaluation questions: Does a guideline differentiate by application risk? Does it state the capacity needed to implement it? Does it offer deployment-level guidance and account for informal adoption? They acknowledge that operationalizing these tiers remains future work. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* ## Institutions - [University of Cambridge](https://md.research.elamur.ai/free/en/universities/university-of-cambridge/) — University, Cambridge, United Kingdom · 256 works · 123,373 citations - [IBM](https://md.research.elamur.ai/free/en/companies/ibm/) — Company, Armonk, United States · 154 works · 72,606 citations - [University of Pittsburgh](https://md.research.elamur.ai/free/en/universities/university-of-pittsburgh/) — University, Pittsburgh, United States · 65 works · 23,401 citations - [University of Michigan](https://md.research.elamur.ai/free/en/universities/university-of-michigan/) — University, Ann Arbor, United States · 86 works · 15,493 citations 📚 Shelves (4): [🤝 Social Sciences](https://md.research.elamur.ai/free/en/shelves/social-sciences/) · 1,954 works · [🤝 Safety Research](https://md.research.elamur.ai/free/en/shelves/social-sciences/safety-research/) · 626 works · [🤝 Sociology and Political Science](https://md.research.elamur.ai/free/en/shelves/social-sciences/sociology-and-political-science/) · 626 works · [🤝 Health](https://md.research.elamur.ai/free/en/shelves/social-sciences/health/) · 6 works 🧭 Topics (1): [🛡️ Bias, fairness & safety](https://md.research.elamur.ai/free/en/topics/bias-fairness-safety/) · 587 works ## Cite Cite as: Nari Yoo, Ashley Khor, Namrata Mukhija, Aminat Adebiyi and Miri Zilka. 2026. Guidelines for Whom? Rethinking AI Ethics in Resource-Constrained Migration Services. Proceedings of the Workshop on Evaluating Evaluations (EvalEval). The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{yoo2026guidelines, title = {{Guidelines for Whom? Rethinking AI Ethics in Resource-Constrained Migration Services}}, author = {Nari Yoo, Ashley Khor, Namrata Mukhija, Aminat Adebiyi and Miri Zilka}, booktitle = {Proceedings of the Workshop on Evaluating Evaluations (EvalEval)}, year = {2026}, doi = {10.18653/v1/2026.evaleval-1.3}, url = {https://doi.org/10.18653/v1/2026.evaleval-1.3} } ``` - [← previous GUI0: Self-Evolving Foundational GUI Agents in Super App Ecosystems](https://md.research.elamur.ai/free/en/papers/wang-2026-gui0-self-evolving-foundational-gui/) - [next → GUNLP at SemEval-2026 Task 10: Emotion-Aware Multi-Task Learning for Conspiracy Detection](https://md.research.elamur.ai/free/en/papers/ziaei-2026-gunlp-semeval-2026-task-10/) - [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/35/) ## References 2 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Tigmanshu Bhatnagar, Maarya Omar, and Davor Orlic. 2025. Bridging AI and humanitarianism: An HCI-informed framework for responsible AI adoption. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems, New York, NY, USA. ACM. Accessed: 2026-3-4. [find](https://scholar.google.com/scholar?q=Tigmanshu+Bhatnagar%2C+Maarya+Omar%2C+and+Davor+Orlic%2E+2025%2E+Bridging+AI+and+humanitarianism%3A+An+HCI%2Dinformed+framework+for+responsible+AI+adoption%2E+In+Proceedings+of+the+2025+CHI+Conference+on+Human+Fact) 2. Johana Bhuiyan. 2023. Lost in AI translation: growing reliance on language apps jeopardizes some asylum applications. [https://www.theguardian.com/us-news/](https://www.theguardian.com/us-news/) 2023/sep/07/asylum-seekers-ai-translation-apps . Accessed: 2026-3-3. \#\# Limitations [↗](https://www.theguardian.com/us-news/) *Updated 2 November 2026*