The rise of generative artificial intelligence is reshaping the human-centered order of knowledge production and academic authorship, rendering the controversy over “AI authorship”a pivotal issue that continuously challenges the ethical boundaries and governance capacity of the academic community. For educational research, the risks triggered by such controversies involve not only legal difficulties in responsibility attribution and accountability, but also a structural deficiency in ethical safeguards centered on learner protection, namely the “guardianship vacuum”.Adopting a genealogical method, this paper firstly traced the historical evolution of authorship and the author function in modern academia, clarifying the institutional origin of authors as the anchor of accountability and their particular significance in educational research. Secondly, based on contemporary disputes, it conducted an institutional analysis of recent cases, academic norms, and governance agendas concerning AI authorship. In this paper, it revealed the deficiencies of the existing knowledge governance frameworks in responsibility chains, transparency, and risk identification, as well as their potential impact on educational research. Finally, building on the above analysis, the paper carried out normative construction and proposed a more rigorous ethical and accountability framework tailored to AI-participated educational research. This study aimed to theoretically and institutionally interpret the authorship disputes in the AI era within the high-risk field of educational research, illuminating its unique ethical dilemmas and providing normative reference for subsequent empirical studies and policy formulation.

As the focus of interdisciplinary attention, the emergent ability of large language models (LLMs) have demonstrated their profound influence on system science, psychology, and linguistics and other fields, and its value for learning science research and practice is also beginning to emerge. Based on this, the paper firstly explained the emergent ability of LLMs from the perspective of learning science, pointed out that LLMs’ emergence ability can not only serve as the subject of learning science, but also provide new means, methods and innovative ideas for the research and practice of learning science, along with new ethical issues. After that, this paper introduced the means and performances of LLMs’ emergent ability to expand the research and practice of learning science, that is, with the help of LLMs’ emergent ability, the expansion of multiple research and practice fields of learning science through prompt engineering, probe, simulation, and content generation and other means. Finally, this paper discussed the research limitations of the current research, and proposed that the future research of learning science needed to continuously and deeply study LLMs’ emergence ability and its influence from three aspects of cognitive model construction, internal mechanism analysis and learning effect evaluation. This paper discussed the interaction relationship between the LLMs’ emergence ability and the research of learning science from multiple angles, and analyzed the promoting effect of the LLMs’ emergence ability on the research of learning science, which provided a new perspective for exploring and understanding the complex process of human learning. In addition, the in-depth analysis of the LLMs’ emergence ability will help us better understand the cognitive mechanism in the learning process, discover new learning models and teaching strategies, and further promote the theoretical and practical innovation of learning science.