AI-Enabled Terrorism: Addressing Radicalization and CBRN Risks in the Age of LLMs

FAR.AI hosted AI-Enabled Terrorism: Addressing Radicalization and CBRN Risks in the Age of LLMs on September 22, 2026 in New York City, on the sidelines of the 81st UN General Assembly (UNGA 81). The event brought together representatives from country missions to the United Nations, UN entities, frontier AI labs, and counter-terrorism practitioners to build a shared understanding of how large language models may enable radicalization, lower barriers to dangerous knowledge, and heighten chemical, biological, radiological, nuclear, and explosives (CBRNE) risks. Following opening remarks from Karl Berzins (FAR.AI) and Steven Siqueira (UN Counter-Terrorism Centre, UNOCT), lightning talks were given by Dina Hussein (OpenAI), David Scharia (UN Counter-Terrorism Executive Directorate), and Brian Tse (Concordia AI). Patricia Paskov (AVERI) then moderated a panel with Adam Gleave (FAR.AI), Adam Hadley (Tech Against Terrorism), and Paul Ash (Christchurch Call Foundation), which looked at technical safeguards, shared evaluation standards, information-sharing infrastructure, and the multilateral coordination needed to reduce these risks.