AI in education, research and administration

• Within education, we encourage the responsible use of AI. Lecturers and programme directors can make use of the AI in Education Hub provided by CELT:

• In the summer of 2025, guidelines were published to support the integration of AI literacy into UT curricula. Implementation is currently underway.

• The University of Twente is also participating in the national pilot of the EduGenAI platform, developed by Npuls, SURF, and higher education institutions. This platform provides access to open-source and commercial language models, as well as computing power through both in-house infrastructure and the public cloud. For more information, you can contact the AI Office.

Strategic Education (SP) is working on a central AI policy for education. For more information, you can contact the AI Office.

UT contributes to the development of AI while critically reflecting on its societal and ethical implications.

We participate in national and European initiatives and are developing a locally hosted large language model (LLM) that aligns more closely with our academic values and privacy requirements.

Strategy & Policy (SP) is working on a central AI policy for research. For more information, you can contact the AI Office.

AI Compliance Framework

UT is developing an AI Compliance Framework to provide guidance on legal and regulatory matters, such as the European AI Act.

As part of this framework, a Decision Tree AI Act has been created to help staff and researchers determine (the download link will be provided soon):
• whether an application qualifies as an AI system;
• whether the AI Act applies;
• the role of the user or provider (developer or deployer); and
• which obligations apply based on the system’s risk level.

Use the AI Act Decision Tree to get an initial indication. Always report your AI application via the AI Office to ensure that the appropriate measures can be taken in line with the AI Act. The Decision Tree is part of the AI Compliance Framework and is intended for internal use within UT. It supports the responsible use of AI in education, research, and operations.

Responsible Use of Generative AI

Until a central licensing model for generative AI becomes available, the university asks all staff and students to handle data, copyrights, and personal information with care.

Cyber Safety and Secure Software Use

The UT applies its existing policies on cyber safety, privacy, and the use of personal hardware and software to all AI applications. This means that when developing or using AI tools or systems, you are expected to follow the same principles that ensure secure and responsible IT use across the university.

Always make sure to use approved and secure software. You can find available and supported applications in the Software Finder

Learn More

More information can be found on the Cyber Safety page, where you can also check whether your AI system complies with the EU AI Act using the AI Act Decision Tree (Checker). 

UT is also exploring the use of AI within its support processes and operational management.

The further development and guidance of these initiatives fall under the AI Compliance Framework, which provides support on matters related to legislation, privacy, and the responsible use of AI.

Before using any new application or tool, please note that, according to the Use of Personal Applications Policy, it is not permitted to accept license agreements that are binding for the university without prior approval. This applies to both paid and unpaid licenses. Please contact Contract Management LISA at for approval and further guidance.

Yes, but do so consciously and responsibly. Never include personal or confidential data, and always verify the accuracy of the output. The UT currently does not have a central licence for tools like ChatGPT or Copilot, so you use them at your own responsibility.

Because no model currently meets all requirements for privacy, security and compliance with the European AI Act. UT is exploring a safe, central licensing option. 

Until then, use AI consciously and responsibly, and avoid sharing confidential or personal data.

Be aware of AI’s possibilities and its risks:

• Protect personal and confidential information.
• Check the reliability and fairness of results.
• Mention when AI was used to create content.
• Follow UT policies on privacy, security and copyright.

That depends on the course. Each programme or lecturer may have their own policy. When in doubt, ask your lecturer. Always use AI transparently and responsibly.

Through the AI in Education Hub and AI literacy guidelines that help embed AI skills and ethics into all curricula.

At the AI in Education Hub (CELT), with best practices, resources and contacts.

UT uses SimCheck for plagiarism checks in Canvas, Remindo and standalone mode.
Detection of AI-generated text is still unreliable; transparency remains key.

Use the Decision Tree AI Act (download link will be provided soon) to check whether your project qualifies as an AI system and what obligations apply. 

Before use always contact AI Office for registration obligations, guidance and help to meet the requirements. Also you can contact the AI Office if you have any questions

Yes, but you remain responsible for privacy, licensing and ethics. Always ensure the terms of use align with UT policies and European law.

A UT-wide framework that helps staff and researchers meet legal, ethical and technical requirements for AI. It includes the Decision Tree AI Act, templates and guidelines

Contact contract management before continuing. They’ll check privacy, security and contractual aspects.

Always assume your input may be stored externally. Never include personal data, student details, research data or confidential material in public AI tools.

Paid licences must go through contract management to ensure legal and security checks.

More information can be found on the Cyber Safety page, which also includes the AI Act Checker and UT policies on cyber safety, privacy, and software use.

Yes, as long as you review the output yourself and comply with UT’s privacy and copyright rules. 

Yes. You may use AI tools to generate images. Just make sure to:

• not use personal data, sensitive information or copyrighted material as input;
• check the settings of the tool to minimise sharing or learning from your prompts;
• clearly indicate when an image was generated with AI.

Tools like Adobe Firefly can be a good option, as they are trained on rights-free materials. You have access to the standard Firefly features (with monthly credits). If you need premium Firefly features or are asked to accept additional licence terms, please contact contract management before proceeding.

The UT is committed to increasing AI literacy — helping students and staff understand how AI works, what its limitations are, and how to use it responsibly in education, research and daily work.

For UT employees and student, several online courses about AI are available on GoodHabitz offering an accessible way to build your knowledge and skills.

Yes. You can use the training request form on this page to suggest a course or workshop. 

Yes, the AI@Work learning community a regional network initiated by the University of Twente, open to all UT employees. Within this community participant-scan learn, experiment, and share experiences about how AI technologies affect work and organisation.

Use the AI Expert Finder to explore researchers, their fields of expertise and publications, and discover how different AI domains are connected.

Yes. If your information is missing or outdated, contact the maintainers listed on the AI Expert Finder page.