Abstract: In the age of intelligence, the convergence and integration of artificial intelligence (AI) with fundamental sciences have emerged as a new frontier for developing scientific research paradigms. This intersection offers promising directions and possibilities for the transformation and advancement of research paradigms in the field of information resource management (IRM). The discipline of IRM must fully harness the permeative, diffusive, and disruptive capabilities of AI technologies, with a focus on insights gained from its role in AI development. This will foster the deep integration of AI with IRM, ultimately constructing a novel research paradigm—what we term the AI+IRM paradigm. First, the generative logic of the AI+IRM paradigm is explored from the perspective of two forms of scientific research paradigms, providing a comprehensive understanding of this new framework. Next, the core concepts of AI+IRM are examined, manifesting as a "double-helix engine", driven by the reciprocal interactions of AI4IRM and IRM4AI. Following this, the shift in research objects under this paradigm transformation is analyzed, including a reconstruction of the information chain model within the AI4IRM paradigm and an exploration of the various levels of AI enhancement under the IRM4AI paradigm. Finally, three paradoxes resulting from the integration of AI and IRM are discussed: the "disciplinary knowledge acquisition—illusion of depth understanding" paradox, the "disciplinary research efficiency—singular disciplinary culture" paradox, and the "disciplinary boundary expansion—weakening of disciplinary ontology" paradox. This paper represents an initial reflection on the transformation of the research paradigm in information resource management in the intelligent era. It offers valuable insights for both the theoretical development and practical innovation within the discipline.