The rapid advancement of artificial intelligence has given rise to a new generation of information systems that no longer function as passive tools but instead operate as autonomous, so-called "agentic" artifacts. These Agentic AI systems are characterized by their ability to independently take on tasks, make decisions under uncertainty, and assume responsibility for complex work outcomes (Baird & Maruping, 2021). Unlike traditional automation solutions, they are capable of autonomously pursuing goals, interacting with other agents, whether human or machine, and dynamically adapting to changing conditions.
This development carries profound implications for the workplace: Traditional notions of IT use, in which humans act as the sole agents steering the system, are increasingly being called into question. Instead, delegation relationships between humans and AI agents are taking center stage—relationships in which tasks, rights, and responsibilities are not unilaterally assigned but dynamically negotiated and distributed (Baird & Maruping, 2021). At the same time, research shows that AI agents are increasingly being perceived and deployed as team members, bringing both opportunities (e.g., reduced conflict when performance is high) and challenges (e.g., lower process satisfaction, lower perceived benevolence) (Dennis, Lakhiwal & Sachdeva, 2023).
Against this backdrop, a fundamental question arises: What will the workplace of the future look like when AI agents are no longer mere tools but active actors collaborating with humans in hybrid constellations? The concept of Hybrid Intelligence, the synergistic combination of human and machine intelligence for solving complex problems, provides a first theoretical lens (Dellermann, Ebel, Söllner & Leimeister, 2019). While AI already outperforms humans in clearly defined tasks, managing ambiguous, context-dependent, and creative tasks in organizational settings still requires human judgment, the question is how these competencies can optimally interact.
References:
Baird, A., & Maruping, L. M. (2021). The next generation of research on IS use: A theoretical framework of delegation to and from agentic IS artifacts. MIS Quarterly, 45(1), 315–341.
Dellermann, D., Ebel, P., Söllner, M., & Leimeister, J. M. (2019). Hybrid intelligence. Business & Information Systems Engineering, 61(5), 637–643.
Dennis, A. R., Lakhiwal, A., & Sachdeva, A. (2023). AI agents as team members: Effects on satisfaction, conflict, trustworthiness, and willingness to work with. Journal of Management Information Systems, 40(2), 307–337.
Organizations face the challenge of managing the transition to Agentic AI not only technologically but also with respect to work design, collaboration, role definitions, and organizational structures. This is not about the macroeconomic question of whether jobs will "disappear," but rather the much more nuanced and practically relevant question of how the nature of work, collaboration, and the work experience at the individual workplace are concretely changing.
The Master thesis may focus on of the following research questions. The final research question will be determined in consultation between the student and the supervisor:
Delegation and Task Allocation Between Humans and AI Agents:
Collaboration and Team Dynamics with AI Agents:
Work Design and Role Transformation:
Trust, Control, and Governance at the Workplace:
Change Processes and Adoption:
Methodology
The methodological approach will be determined in consultation with the supervisor and depending on the chosen research question. Possible approaches include:
Just send a short email to mahei.li@unisg.ch, and I’ll set up a virtual meeting to get to know each other and discuss the details.
I look forward to hearing from you!
Dr. Mahei M. Li