CC BY 4.0Wu, ChanghongMackey, Thomas P.Zhang, WeiruiZou, Jiatong2026-08-172026-08-172026-08-17https://2026.ifla.org/satellite-meetings/https://repository.ifla.org/handle/20.500.14598/7253Generative artificial intelligence (GenAI) has unsettled how information is produced, circulated, and trusted. Deepfakes, synthetic text, algorithmic curation, and model hallucinations now interact in ways the retrieval-and-evaluation model of information literacy was never designed to handle. This paper argues that metaliteracy—an integrative, self-referential framework spanning the behavioural, cognitive, affective, and metacognitive domains—offers a stronger foundation, and that digital resilience is the capability metaliteracy yields when learners face AI-driven disinformation. We report on a meme-based pedagogy taught at the library of Northeast Normal University, a Chinese “Double First-Class” institution, which turns abstract metaliterate objectives into visual, memorable, shareable artefacts. Surveying 183 undergraduate participants, we find consistently high competence across all four domains (means 4.18–4.29 on a five-point scale), strong agreement that memes support recall and reflection, and robust verification behaviour in scenarios involving an unsourced disaster image, a fabricated AI citation, and a deepfake video. We also examine the “dual transformation dilemma” facing Chinese university libraries—reconciling top-down digitalisation mandates with grassroots responses to AI disruption—and how the librarian’s role shifts toward proactive educational designer and steward of information integrity. We offer the model as a scalable, culturally adaptable contribution to information literacy in a disrupted ecosystem. Keywords: metaliteracy; digital resilience; meme-based pedagogy; generative AI; AI-driven disinformationenghttps://creativecommons.org/licenses/by/4.0/Information literacyFake newsArtificial intelligenceArticleWu, ChanghongMackey, Thomas P.Zhang, WeiruiZou, Jiatong