AI-mediated IS development
Track 15
Track chairs
Nikolaus Obwegeser
Bern University of Applied Sciences
Noel Carroll
University of Galway
Anita Neumannova
Vienna University of Economics and Business
Track description
Information systems (IS) development is undergoing a fundamental transformation. The rapid diffusion of AI-mediated software development, ranging from in-editor assistants and agentic coding tools to dedicated platforms that generate software directly from natural-language intent, is reshaping who writes code, how it is produced, and how its quality is assured. The most pointed expression of this shift is vibe coding, a paradigm in which developers describe what they want in natural language and let large language models generate the code, deliberately refraining from inspecting or modifying it. AI-mediated development opens many new questions around the delegation of cognitive labour to probabilistic systems, deliberate code opacity, and the displacement of traditional control mechanisms.
AI-mediated development is reshaping hiring patterns, educational curricula, and the economics of software production. Early evidence points to fundamental shifts in how organisations allocate engineering resources, how developers experience and structure their work, and how software enters production. Further, existing theoretical frameworks are strained by the novelty of the phenomenon. Classical assumptions about code authorship, developer expertise, peer review, and accountability were built for a world in which humans wrote code they could understand. When that assumption breaks down, the conceptual foundations of IS development governance need rebuilding. In addition, the gap between practice and research is widening. Industry adoption has moved faster than empirical investigation, and the discourse on AI-mediated development is dominated by technical and commercial voices rather than rigorous analysis from IS scholars.
IS research has long examined the sociotechnical consequences of shifting the boundary between human and machine work, and offers mature theoretical resources for doing so. This track, therefore, invites research that engages critically and empirically with AI-mediated IS development across its full spectrum, from AI pair programming through agentic tools to vibe coding. The scope extends beyond tool adoption to address how organisations govern opaque artefacts, how developers calibrate trust in probabilistic code generation, how accountability functions when neither human nor AI can fully explain implementation decisions, and how this paradigm reshapes the division of cognitive labour.
The track aligns with the ECIS 2027 theme “Bridging Digital Borders” by examining how AI-mediated development crosses and redraws the boundaries between developers and non-developers, between humans and autonomous systems, and between regulatory regimes that govern digital work. We welcome design-science, behavioural, qualitative, quantitative, and conceptual contributions, and particularly encourage work that connects practitioner evidence with established IS theory.
Topics of interest
- Empirical studies of AI-mediated coding adoption, use patterns, and outcomes in organisations
- Impact of AI-mediated development on organisational strategy and practices
- Changing roles and identities of software developers
- Trust calibration and mental models in human-AI code co-creation
- Governance and control mechanisms for opaque AI-generated code (including prompt-level guardrails, trace-based auditing, and portfolio-level gate criteria)
- Accountability, liability, and provenance when code is co-produced by humans and probabilistic AI systems
- Socio-technical reconfiguration of development work
- Developer motivation, job satisfaction, and skill evolution under AI-mediated development
- Security and quality risks of AI-generated code, including empirical studies of vulnerability patterns
- Shadow AI and the democratisation of software creation in enterprises
- Transitions and governance between development modes (e.g., vibe coding, low-code/no-code, and traditional)
- Interaction traces as documentation, design rationale, and organisational knowledge
- Regulatory and ethical implications under frameworks such as the EU AI Act and GDPR
- Design-science artefacts that support responsible vibe coding (trace repositories, prompt-governance systems, integrated controls)
- Theory development for AI-mediated IS development
Associate Editors
Margeret Hall
University of Nebraska at Omaha, USA
Everist Limaj
Vienna University of Economics and Business, Austria
Edona Elshan
Vrije University Amsterdam, The Netherlands
Jennifer Hehn
Bern University of Applied Sciences, Switzerland
Lise Tordrup Heeager
Aarhus University, Denmark
Anna Wiedemann
Bern University of Applied Sciences, Switzerland
Roman Rietsche
Bern University of Applied Sciences, Switzerland
Keld Pedersen
Aarhus University, Denmark
Daniel Burkhardt
Ferdinand Steinbeis Institut, Germany
Stephen McCarthy
University College Cork, Ireland
Till J. Winkler
University of Hagen, Germany
Andreas Hein
University of St.Gallen, Switzerland
Matthew Ajimati
University of Galway, Ireland
Herath Pathirannehelage Savindu
ETH Zürich, Switzerland
Tomoko Yokoi
University of St. Gallen, Switzerland
Domenico di Prisco
IESEG School of Management, France
