UN warns AI amplifies misinformation targeting refugees

The UN's warning reflects a widening recognition that AI tools, while neutral technology, can be weaponized to harm vulnerable populations at unprecedented scale. UNHCR identified a pattern where conflict and displacement often occur alongside what it calls "information crises" — the deliberate or accidental spread of false narratives that dehumanize refugees and create the conditions for violence. Smugglers and traffickers are also exploiting digital platforms to deceive people with false promises of safety and employment, pushing them into dangerous situations.

The problem is not new — misinformation has long fueled xenophobia and violence — but generative AI has made it worse in two ways. First, deepfakes and automated content generation allow false narratives to spread faster and at greater scale than manual creation would permit. Second, content moderation tools designed for major languages often fail in the smaller-language contexts where many refugee crises unfold, leaving vulnerable populations exposed. UNHCR's emphasis on "less-common languages" underscores that even well-intentioned platform safeguards can inadvertently create gaps where harm flourishes. The agency's call for tech companies to "partner" and "collaborate" signals that humanitarian organizations alone cannot solve the problem — the tools and platforms that amplify speech must be redesigned to serve humanitarian protection, not just commercial interests.