<!-- TITLE: Evaluating GenAI Innovation in Higher Education. A Whitepaper -->
<!-- DESCRIPTION: A five-stage framework for planning, piloting and evaluating generative AI in higher education teaching — derived from projects at TU Graz. -->
<!-- LANG: en -->
<!-- URL: https://flaait.tugraz.at/en/publications/2026/schoen-2026-evaluating-genai-innovation/ -->
<!-- URL_DE: https://flaait.tugraz.at/publikationen/2026/schoen-2026-evaluating-genai-innovation/ -->
<!-- URL_EN: https://flaait.tugraz.at/en/publications/2026/schoen-2026-evaluating-genai-innovation/ -->
<!-- TYPE: publication -->
<!-- PUBLISHED: 2026-05-14 -->
<!-- YEAR: 2026 -->
<!-- PUBLICATION_TYPE: report -->
<!-- VENUE: Whitepaper -->
<!-- PUBLISHER: Graz University of Technology -->
<!-- AUTHORS: Schön, Brünner, Ebner, Leitner -->
<!-- PUBLICATION_LANGUAGE: en -->
<!-- DOI: 10.3217/1tfkj-ntj85 -->
<!-- OPEN_ACCESS: true -->
<!-- PILLARS: ai -->
<!-- PARTNERS: tugraz_tu-graz, tugraz_hcc -->

# Evaluating GenAI Innovation in Higher Education. A Whitepaper

*A five-stage framework for planning, piloting and evaluating generative AI in higher education teaching — derived from projects at TU Graz.*

*Schön, S., Brünner, B., Ebner, M., & Leitner, P. (2026). Evaluating GenAI Innovation in Higher Education. A Whitepaper. Whitepaper. Graz University of Technology. https://doi.org/10.3217/1tfkj-ntj85*

**Key facts**

- **Type of publication:** Report
- **Authors:** Schön, S., Brünner, B., Ebner, M., & Leitner, P.
- **Year:** 2026
- **Published in:** Whitepaper
- **Publisher:** Graz University of Technology
- **DOI:** [10.3217/1tfkj-ntj85](https://doi.org/10.3217/1tfkj-ntj85)
- **Open Access:** Yes

The whitepaper proposes a non-linear five-stage evaluation framework: specifying context and use cases, assessing legal, ethical and technical feasibility, selecting an implementation strategy, piloting with multi-layered evaluation and carrying out data-informed analysis.

It is based on a cross-case synthesis of AI applications at TU Graz – including AI-generated multilingual educational videos, RAG-based chatbots and automated assessment. Generative AI is understood as a pedagogical intervention that requires continuous governance, human oversight and institutional accountability: applications are refined, scaled or discontinued based on evidence of their educational value.

## Participating partners

- [TU Graz](https://flaait.tugraz.at/en/partners/tu-graz/)
- [Institute of Human-Centred Computing](https://flaait.tugraz.at/en/partners/hcc/)
