The AI4Debunk consortium successfully convened for its latest General Assembly from April 9 to 10, 2026, in the historic city of Rome. Hosted by CNIT (Consorzio Nazionale Interuniversitario per le Telecomunicazioni) and CNR-IRPPS, the two-day event marked a pivotal mid-project milestone focused on transitioning from foundational research to deployable, trustworthy AI solutions.
Mini-Symposium: AI and Disinformation
The assembly opened with a high-level Mini-Symposium, bringing together experts to bridge the gap between technical innovation and societal impact. The guest speakers provided the momentum needed for the consortium’s deep-dive sessions:
Dr. Mario Paolucci (IRPPS-CNR): Explored the conceptual roles of AI in “Ucronia” (alternate histories) and how these narratives fuel modern disinformation.
Prof. Tiziana Catarci (Sapienza/SIpEIA): Critically examined the “Human-In-The-Loop” approach, advocating for ethical AI frameworks that truly empower users.
Prof. Dr. Fabrizio Silvestri (Sapienza): Detailed the “front lines” of Adversarial Machine Learning, focusing on how AI can be trained to resist manipulation during content moderation.
Dr. Francesco Saverio Nucci: Highlighted media literacy as the “ultimate shield,” emphasizing that technology must be paired with human education.
Dr. Andrea Nicolai (T6 Ecosystems): Reflected on EU strategies and the essential role of IDMO/EDMO hubs in creating a unified front.
The morning concluded with a cross-disciplinary panel featuring Kevin El Haddad, Jan Kragt, and Inna Šteinbuka, who discussed the urgent need to synchronize technology with policy and education.
Technological Progress & Roadmap
Led by Project Coordinator Prof. Dr. Inna Šteinbuka, the consortium reviewed key technical milestones and set a path for 2026–2027:
Progress was reported on Knowledge Graphs, a new disinformation scoring system, and a robust debunking API.
Partners showcased the development of browser plugins and AR/VR environments designed to make fact-checking intuitive.
A major focus remains on explainability, ensuring the AI’s debunking process is transparent and unbiased.
The Rome GA concluded with a clear mission: delivering a powerful, multidisciplinary toolkit that empowers citizens and policymakers alike in the fight against digital falsehoods.
