Generative AI is increasingly transforming how open educational materials are created. A twillo survey conducted in spring 2025 shows how higher education institutions, instructors, and active users are already applying AI, where they see potential, and where uncertainties remain. This article summarises key findings and contextualises them for OER practice.
Purpose and Structure of the twillo Survey
Between April and June 2025, twillo conducted a mixed-methods study on the use of generative AI applications for creating and disseminating OER. Three groups were surveyed:
• Early adopters from Germany and Austria – individuals already using AI in OER creation
• Vice presidents for teaching and programme directors at universities in Lower Saxony
• active twillo users
The aim of the study was to identify existing areas of application, potentials, and challenges associated with using AI technologies and tools for OER creation.
High Potential: How AI Supports OER Creation
The results paint a clear picture: Responding institutions and individuals are largely open to the possibilities of AI and are already using a wide range of tools. At the same time, respondents emphasise that clear boundaries and responsibilities are essential when integrating AI.
61.7% of the 102 respondents attribute significant potential to the use of AI in creating and further developing OER. AI primarily supports instructors by generating visualisations, optimising texts for different target groups, translating materials into numerous languages, and thereby enabling international use of OER. Several respondents also highlight that AI can contribute to enhancing accessibility, for instance through suggestions for alternative texts.
Practical guidance on creating educational materials can be found in the section “AI in open university teaching”
Challenges: Ensuring Quality, Legal Clarity, and Didactic Integrity
Despite these potentials, the use of AI is not without risks. Respondents frequently mention quality issues such as hallucinations or distortions that may undermine the reliability of results. Legal uncertainties related to copyright and licensing also remain a major concern for universities, increasing the need for clear guidelines.
Additionally, some respondents fear that didactic expertise might become overshadowed. Universities therefore need to find ways to position AI as a supportive rather than substitutive element in the pedagogical design process. The rapid evolution of technology presents further organisational challenges, as tools and quality standards are continuously changing. Finally, respondents point out that producing high-quality OER with AI still requires time — for example, through precise prompting and careful review of outputs.
Further insights into the impact of generative AI on teaching and learning are provided in the working paper “How AI Is Transforming Higher Education”, which highlights key developments and areas for institutional action.
Conclusion and Outlook
The twillo survey demonstrates that AI applications can make a significant contribution to the creation, refinement, and dissemination of OER. Visualisation, text optimisation, translation, and metadata generation offer tangible benefits for educators and higher education institutions. At the same time, a reflective approach to the limitations of AI systems remains essential — including quality assurance, didactic design considerations, and legal frameworks. When these factors are carefully combined, AI can meaningfully support and advance open higher education.
The full study results and resulting conclusions are expected to be available in early 2026 via the HIS-HE media centre.
Stay Up to Date
If you don’t want to miss upcoming studies, events, and practical insights related to OER and open higher education, we recommend subscribing to our newsletter. Subscribe to the twillo newsletter in german language.