Paper Type
Complete
Paper Number
PACIS2026-1577
Description
Knowledge and technology transfer units in German public non-university research institutions increasingly experiment with AI, especially large language models, yet adoption remains fragmented and rarely aligned with transfer workstreams. This paper presents a comparative AI use case matrix based on 12 semi-structured expert interviews across major German public research institutions. Using an inductive approach, we distinguish AI use cases already in use, including pilots, from those framed as desired or high-potential, and structure them along a modular KTT reference model from research to spin-off management. The results show a small set of broadly adopted entry-level applications for language-centric knowledge work and content preparation, alongside a larger portfolio of aspirational applications targeting process support, information discovery and triage, evidence synthesis, and decision-oriented workflow automation. We consolidate these findings in a comparative matrix and derive implications for KTT practice, AI-enabled process augmentation, and future evaluation.
Recommended Citation
Lovrekovic, Josip; Tavakoligargari, Masoumeh; Etzkorn, Manuel; Gieß, Anna; Jürjens, Jan; von Korflesch, Harald F. O.; and Fettke, Peter, "AI Use Cases in Knowledge and Technology Transfer: Evidence from Expert Interviews" (2026). PACIS 2026 Proceedings. 6.
https://aisel.aisnet.org/pacis2026/di_entren/di_entren/6
AI Use Cases in Knowledge and Technology Transfer: Evidence from Expert Interviews
Knowledge and technology transfer units in German public non-university research institutions increasingly experiment with AI, especially large language models, yet adoption remains fragmented and rarely aligned with transfer workstreams. This paper presents a comparative AI use case matrix based on 12 semi-structured expert interviews across major German public research institutions. Using an inductive approach, we distinguish AI use cases already in use, including pilots, from those framed as desired or high-potential, and structure them along a modular KTT reference model from research to spin-off management. The results show a small set of broadly adopted entry-level applications for language-centric knowledge work and content preparation, alongside a larger portfolio of aspirational applications targeting process support, information discovery and triage, evidence synthesis, and decision-oriented workflow automation. We consolidate these findings in a comparative matrix and derive implications for KTT practice, AI-enabled process augmentation, and future evaluation.
Comments
09-Transformation