Adaptation in Agentic AI: A Structured Literature Review

Agentic AI Systems, defined as being goal-directed, multi-step, and tool-using, are being adopted at an increasingly rapid pace across all industries. Within this shift, the concept of “adaptation”, i.e., how these systems adjust to new environments, changing user intents, and novel constraints, is the centerpiece of not just system performance and strategic value, but also governance, safety, and auditability. Consequently, adaptation is at the forefront of Information Systems (IS) discussions. However, current literature, particularly in non-technical and IS-facing venues, uses the term inconsistently.

We are looking for motivated students to systematically review and theorize the concept of adaptation in agentic AI. The student will utilize the PRISMA framework and a mixed-methods approach that encompasses both academic databases and grey literature to clarify how adaptation is conceptualized across practical and scientific domains and propose a unifying taxonomy. Contributions are intended to be accessible to a broad audience, featuring clear exposition, a lively tutorial style, and pointers to the literature for further study.

For more information please contact Wenyuan "Rafael" Wu