Civilian AI Exposes Governance Gaps in Post-Conflict Settings
A Just Security analysis published July 1, 2026, by Mariana Beselga argues that civilian AI systems administering rights and services face far less international governance attention than military AI, despite a 2024 warning from the U.N. Secretary-General's High-level Advisory Body on Artificial Intelligence of a "systemic accountability deficit" in civilian deployments. The piece uses Colombia's Sisben IV social-benefit classification system and Ukraine's Diia digital-government platform as case studies, noting a Global Index on Responsible AI finding that 67% of 138 countries scored 25 points or less out of 100 on responsible-AI governance. The author, founder of the Policy Hub for Inclusive Development, argues this gap is most consequential in fragile, post-conflict settings where institutional capacity and public trust are already weak.
The practical takeaway for anyone building or procuring public-sector AI in fragile states: opacity that might be a manageable UX or compliance issue in a high-capacity democracy becomes a legitimacy risk when it touches a population's access to welfare, identity, or peace-agreement benefits, and post-conflict contexts multiply that risk because institutional appeals processes are often the first thing to erode.
What happened
Just Security published an analysis by Mariana Beselga, founder and co-CEO of the Policy Hub for Inclusive Development, arguing that international AI governance has concentrated on military applications, particularly autonomous weapons systems addressed in U.N. General Assembly resolutions 79/239 and 80/58, while civilian AI systems that mediate access to social protection, identity, and services have received comparatively little governance attention. The article cites the U.N. Secretary-General's High-level Advisory Body on Artificial Intelligence, which warned in its 2024 "Governing AI for Humanity" report of a "systemic accountability deficit" in civilian AI deployment, cautioning that automated decision-making is advancing faster than oversight mechanisms can absorb.
Case studies cited
The piece examines Colombia's Sisben system, the country's primary infrastructure for allocating social assistance since the 2016 peace agreement. Sisben evolved from household-survey-based assessments into Sisben IV, which uses predictive algorithms and administrative data, including credit histories and telecommunications data, to produce "probabilistic" socioeconomic classifications; the author argues this shift toward inferential scoring has made eligibility decisions harder for beneficiaries to contest, even though formal appeals mechanisms remain available. Ukraine's Diia digital-government platform is cited as a contrasting case, showing both the value of digital infrastructure in sustaining public services during wartime and the risks, around data integrity, interoperability, and public trust, when core state functions depend on digital systems under stress. The author also cites the Global Index on Responsible AI, which assessed 138 countries and found 67% scored 25 points or less out of 100 on responsible AI governance, meaning roughly six billion people lack robust human-rights and accountability protections under current AI oversight regimes.
Editorial analysis
This is an opinion and policy-analysis piece, not breaking news; its core argument, that governance norms developed for military AI (human oversight, traceability, accountability) have not been extended to civilian systems with comparable urgency, is a reasonable synthesis of existing multilateral reporting (the UN Advisory Body report, OECD public-sector AI research) rather than a novel empirical finding. The Colombia and Ukraine cases are used illustratively; readers should treat the framing as the author's normative argument rather than a comprehensive audit of either country's system.
What to watch
Whether donor and multilateral procurement rules begin requiring transparency and appeal mechanisms for AI-mediated public-benefit systems, whether the OECD or UN advisory bodies extend military-AI-style oversight norms to civilian deployments, and how post-conflict digital-identity systems like Diia address contestability as they scale.
Key Points
- 1International AI governance has developed detailed accountability norms for military autonomous weapons but has not extended similar oversight to civilian AI administering welfare and services.
- 2Colombia's Sisben IV uses predictive algorithms and administrative data to classify welfare eligibility, a shift the author says makes eligibility decisions harder for citizens to contest.
- 3A Global Index on Responsible AI assessment found 67% of 138 countries score poorly on AI governance, leaving billions without robust accountability protections as public-sector AI use grows.
Scoring Rationale
A substantive, well-sourced policy analysis on a genuinely under-examined governance gap, grounded in the U.N. Advisory Body's 2024 report and concrete case studies (Colombia's Sisben IV, Ukraine's Diia). Raised slightly from 4.5: the prior rationale incorrectly flagged the primary source as off-topic (confusing it with an unrelated Just Security 'Read Next' link); the actual Just Security article is on-topic, single-authored opinion/analysis rather than breaking news, which caps it in the minor-to-solid range.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
