Aim and Topics
17 November 2026Overview
Now in its seventh edition, the ROMCIR Workshop focuses on the development and evaluation of information access systems aimed at mitigating information disorder, including both human-generated and AI-generated misinformation and disinformation, across different domains and online environments.
By information disorder, we refer broadly to different forms of information pollution, ranging from unintentionally inaccurate or misleading content to deliberately fabricated or manipulated information, whether produced or disseminated by humans or automated systems.
ROMCIR welcomes research addressing the factual accuracy, credibility, reliability, and trustworthiness of online information. Its scope encompasses different types of content (e.g., Web pages, news, reviews, medical information, social media posts, and AI-generated content), platforms (e.g., search engines, social networking services, question-answering systems, and conversational AI interfaces), and tasks (e.g., misinformation detection, fact-checking, credibility assessment, trustworthy information retrieval, and information verification).
The workshop aims to bring together researchers and practitioners from Information Retrieval, Natural Language Processing, Artificial Intelligence, Social Computing, and related disciplines to discuss methods, applications, evaluation frameworks, and emerging challenges in providing reliable access to information.
SPECIAL FOCUS FOR 2027: TRUSTWORTHINESS IN GENERATIVE AND AI-MEDIATED INFORMATION ACCESS
This edition places special emphasis on a major shift reshaping how users access information: increasingly, information access systems provide synthesized answers, often accompanied by citations, rather than traditional ranked lists of documents.
Generative search engines, retrieval-augmented generation (RAG) systems, conversational assistants, and deep research agents offer new opportunities for accessing and understanding information. At the same time, they introduce challenges concerning factual accuracy, source attribution, evidence verification, transparency, and user trust.
Recent studies auditing AI-generated search summaries and deep research systems have highlighted limitations in citation accuracy, the presence of unsupported claims, and the potential for incomplete or one-sided answers. Moreover, emerging adversarial practices, including knowledge poisoning and manipulation of generative search pipelines, raise further concerns about the integrity of AI-mediated information access. Regulatory developments, such as the European Union’s Digital Services Act (DSA) and AI Act, also reinforce the importance of transparency, accountability, and trustworthy information ecosystems.
Building on previous ROMCIR editions, the 2027 workshop particularly encourages contributions investigating how information retrieval and related techniques can support the factual reliability, verifiability, transparency, and robustness of AI-generated answers.
This special focus complements, rather than restricts, the workshop’s broader scope. Contributions addressing traditional and emerging forms of information disorder, credibility assessment, misinformation detection, and trustworthy information access remain equally welcome.
Topics of Interest
The topics of interest include, but are not limited to, the following:
- Access to and retrieval of reliable and truthful information
- Adversarial manipulation of generative search (e.g., generative engine optimization, knowledge poisoning)
- Agentic AI and deep research agents for information generation, dissemination, and truthfulness assessment
- Attribution, citation faithfulness, and verifiability in generative search and RAG
- Auditing of AI-mediated information access (e.g., chatbots, AI Overviews, deep research agents)
- Bot, spam, and troll detection
- Computational fact-checking, including multimodal approaches
- Credibility assessment of online documents and sources
- Crowdsourcing for information truthfulness assessment
- Disinformation, misinformation, and bias detection
- Evaluation strategies, benchmarks, and datasets for information truthfulness
- Explainability and transparency of search results and generated answers
- Generative models and information truthfulness assessment
- Human-in-the-loop misinformation detection and user trust in synthesized answers
- Information polarization in online communities and echo chambers
- Multilingual and cross-lingual misinformation detection, fact-checking, and trustworthy information access, including low-resource languages
- Propaganda identification and analysis
- Query reformulation strategies for truthful information retrieval
- Regulation, governance, and policy for trustworthy information and AI (e.g., DSA and AI Act transparency obligations)
- Retrieval-augmented generation with conflicting, incomplete, or low-quality evidence
- Search patterns and simulation strategies for information verification
- Security, privacy, and information truthfulness
- Societal reactions to misinformation
- Stance detection
- Trust and reputation
Data-driven approaches, particularly those supported by publicly available datasets, reproducible experimental settings, and transparent evaluation methodologies, are especially welcome.