<![CDATA[This study examines how graduate students in Library and Information Science (LIS) in Indonesia experience generative artificial intelligence (AI) in their academic work and how disciplinary and professional backgrounds shape that experience. Using Interpretative Phenomenological Analysis (IPA), six LIS master's students who regularly used AI for academic purposes took part in semi-structured interviews. Analysis involved initial noting, case-level theme development, within-case interpretation, and cross-case comparison. Four themes were developed: AI as epistemic infrastructure; efforts to preserve cognitive autonomy amid dependency; AI as both a cognitive and affective space; and the uneven influence of LIS-related professional dispositions.Two recurring patterns were also identified. Epistemic hierarchy reversal describes participants' tendency to consult AI before conventional sources, while evaluative offloading refers to cases in which AI became a reference point for judging one's own ideas. The interviews were consistent with cognitive offloading and, in some cases, possible cognitive coupling, but participants also described deliberate efforts to protect authorship, independent judgment, and cognitive boundaries. LIS-related habits appeared in verification, citation, information organization, and ethical reflection, although not uniformly. The study adds an idiographic account of how generative AI can become woven into academic thinking without assuming that routine use is equivalent to extended cognition. It also points to practical questions for LIS education around information evaluation, authorship, and ethical use of AI]]>
