Articles

Last updated: 2026-08-15

Question-led articles on applied AI: LLM security in public institutions, AI chatbots for crisis-affected people, risk registers and post-launch review, why government AI projects fail, AI governance gaps in the Middle East, running AI in low-connectivity settings, data minimisation, and gender-based violence case management.

Each article answers a specific question that recurs in practice, drawn from delivering AI systems inside United Nations operations. They are written to be read by someone with a decision to make.

  • AI Chatbots for Crisis-Affected People
    When an AI chatbot for refugees or crisis-affected people is safe to deploy, when it is not, and the design conditions that separate the two.
    6 min · updated 2026-08-15
  • AI and Gender-Based Violence Case Management
    Where AI can and cannot support gender-based violence case management, and why this is among the highest-risk applications in humanitarian work.
    6 min · updated 2026-08-15
  • AI Governance Gaps in the Middle East
    Where AI governance in the Middle East is thin, how Gulf and Levant governments face different problems, and what inclusive AI governance means for the region.
    7 min · updated 2026-08-15
  • Running AI Where the Network Is Not
    How to run AI in low-connectivity and offline settings, and why systems that pass every laboratory test fail in field conditions.
    6 min · updated 2026-08-15
  • The AI Risk Register and What Happens After Launch
    How to maintain an AI risk register, how often AI systems should be reviewed after deployment, and what responsible AI looks like day to day inside a UN agency.
    6 min · updated 2026-08-15
  • Data Minimisation Against AI Value
    How to resolve the tension between collecting less personal data and building useful AI: decide what the system must accomplish before deciding what it may collect.
    6 min · updated 2026-08-15
  • LLM Security in Public Institutions
    How to manage LLM security risk in a public institution: the staff data-leakage problem, the model as an untrusted input path, and when on-premises deployment is justified.
    7 min · updated 2026-08-15
  • Why Government AI Projects Fail
    The recurring causes of failed public-sector AI projects, and the portfolio discipline that prevents them: knowing what you run before deciding what to build.
    7 min · updated 2026-08-15