INLG 2026 — Natural Language Generation, Utrecht

INLG 2026 is the 19th International Conference on Natural Language Generation, held October 17-21, 2026 in Utrecht, the Netherlands, and timed to sit just before EMNLP 2026 so attendees can chain the two trips. Organized under SIGGEN (the ACL special interest group on natural language generation) and hosted by Utrecht University's NLP group, INLG is the dedicated home of NLG research - the subfield concerned with how systems produce fluent, faithful, controllable text. That focus has never been more central: the generative turn that put large language models at the heart of AI is, at its core, a natural-language-generation problem, and INLG is where the people who study generation rigorously - faithfulness and hallucination, controllability, evaluation methodology, data-to-text, summarization, dialogue generation, and the perennially hard question of how to measure generated text - convene. For attendees, the appeal is depth over breadth: where a giant venue spreads NLG across hundreds of unrelated papers, INLG concentrates the community, making it an unusually good place to get detailed feedback, find collaborators, and track where generation evaluation is actually heading.

The conference accepts regular research papers and system-demonstration papers, all peer-reviewed and archived in the ACL Anthology. INLG 2026 offers two submission routes: direct submission to the conference (regular and demo papers due July 15, 2026, AoE), and a commitment route via the ACL Rolling Review for authors whose papers were reviewed in an ARR cycle (ARR commitment to INLG due August 5, 2026). A non-archival track (deadline August 22) lets researchers present work intended for publication elsewhere or already published, which suits cross-posting and work-in-progress. Notifications land August 15 and camera-ready papers are due September 7.

The program is rounded out by workshops and tutorials, so the final days typically host focused satellite events on emerging NLG themes. Who should attend: NLP and computational-linguistics researchers specializing in generation; PhD students working on text generation, summarization, dialogue or evaluation; and industry research scientists building products on top of generative models who want grounding in the methodological state of the art, especially around measuring quality and curbing hallucination. Because it precedes EMNLP and sits in a compact, well-connected European city, INLG 2026 is also an efficient stop for researchers planning a single autumn conference trip. With generation now the dominant mode of AI interaction, a venue devoted entirely to doing it well - and evaluating it honestly - is more relevant than its modest size suggests, and INLG remains the field's anchor meeting for that work.