Full Breakdown
UNHCR Warns AI-Powered Misinformation Endangers Refugees and Aid Workers
7/8/2026, 8:42:01 AM
AI-Driven Misinformation Threatens Displaced Populations
UNHCR officials warned on 7 July that misinformation, hate speech and AI-generated deepfakes are increasingly linked to physical harm against refugees and humanitarian personnel. The agency highlighted that distorted information can impede access to protection services, education and employment, and can fuel hostility in host communities.
Information Crises Behind Major Displacement
UNHCR describes the world’s largest displacement emergencies as “information crises” as well as humanitarian ones. The agency notes that false narratives often spread alongside conflict-driven movements, amplifying risks for people fleeing persecution. In Libya, for example, AI-generated videos impersonating UNHCR officials and online calls to reveal staff locations have heightened tensions and endangered aid operations.
Scale of the Problem
- A UNHCR staff survey found that 93 % of respondents had encountered misinformation, disinformation or hate speech affecting their work.
- At the end of 2025, 117.8 million people were forcibly displaced worldwide, including 68.7 million internally displaced persons and 35.6 million refugees under UNHCR’s mandate.
- Two-thirds of those refugees originate from Venezuela, Ukraine, Syria, Afghanistan and Sudan.
- The Information Integrity Response Toolkit, developed with local partners, has been used by roughly 2,000 practitioners for risk assessments and community-based responses.
UNHCR’s Call to Action
UNHCR urged technology firms, governments and researchers to embed humanitarian concerns in emerging AI governance frameworks. The agency called for improved content-moderation tools that operate effectively in less-common languages and for the inclusion of refugee perspectives in AI standards discussions. Initiatives cited include the Community of Practice on Information Integrity, the Global Refugee Forum Multistakeholder Pledge launched with Google in 2023, and ongoing collaborations at the AI for Good Summit in Geneva.
Verbatim Quotes
- “The spread of misinformation, hate speech and deepfakes is exacerbating and inciting real-world harm to refugees and humanitarians,” — Gisella Lomax, UNHCR senior adviser on information integrity
- “When information is distorted, it can reduce access to jobs and education, make integration harder and threaten social cohesion.” — UNHCR statement, AI for Good Summit
- “Women – among both refugees and staff – are being disproportionately targeted.” — UNHCR briefing note
- “This isn’t about how legitimate societal concerns on migration and asylum are debated online – but about life-threatening information risks that can lead to serious harm in humanitarian contexts.” — UNHCR senior official, briefing note
What’s Next
UNHCR plans to expand the Information Integrity Response Toolkit, strengthen the Community of Practice, and pursue further multistakeholder pledges on AI governance. The agency will continue to monitor AI-generated content and advocate for policy measures that protect displaced populations and humanitarian workers.
