<![CDATA[Generative artificial intelligence changes the production and circulation of terrorist propaganda while dispersing human involvement across content creation, editing, and dissemination. This article examines the boundary between artificial-intelligence-assisted terrorist propaganda, incitement to terrorist offences, and protected expression in Indonesia, and develops a proportionate attribution framework through comparison with European Union law. It employs normative juridical research with statutory, conceptual, and comparative approaches and descriptive-prescriptive analysis. Article 13A of the Indonesian counter-terrorism legislation is assessed alongside general criminal-law principles, Directive (EU) 2017/541, and Regulation (EU) 2021/784. A 2026 CTED briefing on Tehrik-e Taliban Pakistan is used only as a factual illustration, not as adjudicated proof. The study finds that Article 13A remains applicable because liability attaches to human conduct, but organisational nexus, intent, and individual contribution require separate proof. European Union law offers a clearer distinction between individual criminal liability and platform content-governance duties. The article proposes five stages: identifying content, classifying conduct, testing offence elements, examining context and rights, and selecting a proportionate response. Interpretive guidance and explicit safeguards for legitimate expression are prioritised over immediate expansion of criminal offences.]]>
