Approved by the Executive Board, the document offers faculties guidance for developing their own policies and outlines both the opportunities and risks of this rapidly evolving technology.
Read the key points below or consult the full document.
Why a framework for genAI in education?
Generative AI and its rapid development are already impacting society, and higher education is no exception. AI can support learning, for instance in writing or processing information. At the same time, it brings risks: from the erosion of academic skills to data privacy concerns, the spread of misinformation, and increased environmental pressure.
VU Amsterdam takes responsibility by setting clear guidelines aimed at a safe and future-oriented learning environment. The framework will be regularly updated.
AI should support academic development
GenAI is not a replacement for independent thinking, writing, or research. Students learn to use AI responsibly to support their learning, not to undermine it. Working independently, thinking critically, and distinguishing fact from fiction remain key skills.
AI literacy is a core skill
Every student and lecturer develops a basic understanding of AI:
- How does AI work, and what are its limitations?
- When is using AI helpful or unsuitable?
- What are the ethical issues involved (such as bias, copyright, and sustainability)?
Lecturers play an active role in developing students’ AI literacy in their courses and are encouraged to follow training where needed.
Scientific integrity is the foundation
AI use in education and research must always align with the principles of scientific integrity. Transparency about AI use and critical reflection on AI-generated output are essential.
Sustainability and ethics matter
AI has a significant environmental footprint due to the energy consumption of large language models. It’s also important to critically assess how AI tools are developed and trained. Students and lecturers are encouraged to use AI consciously and sustainably.
Clarity in courses and assessments
For each course, it must be clear:
- Whether the use of AI is allowed and in what form.
- What rules apply to assignments and assessments.
- How students should report their AI use.
If not explicitly stated otherwise, students should assume that original work should be made without AI assistance.
Privacy and data security are essential
Students cannot be required to use AI tools for which VU Amsterdam does not have a license. Privacy, ethics, and software transparency are key conditions.
No personal data or confidential (research) information may be processed through AI tools. Students, lecturers, and staff should be aware of data collection risks by AI providers.
Fraud prevention through formative guidance
AI fraud is difficult to detect. AI detectors are unreliable and must not be used. Instead, formative assessment is a better approach. Regular feedback moments help monitor students’ development and allow early identification of issues.
Pilots and innovation under clear conditions
Experimenting with AI in education is possible, but only under certain conditions:
- Report the pilot in advance to the VU Education Lab.
- Plan carefully and assess the risks.
- Evaluate and share results for wider application.
Support and shared responsibility
Everyone at VU Amsterdam – the Executive Board, faculties, examination boards, lecturers, staff, and students – plays a part in using generative AI in thoughtful and appropriate ways. Together, thanks to strong collaboration and resources, we are shaping a learning environment that is prepared for the responsible use of generative AI.
Support is available via the VU Education Lab for pilots and knowledge sharing, and via the new VU Artificial Intelligence Competence Network, which offers technical, legal, and didactic support. Lecturers and teaching staff can follow training via the Centre for Teaching and Learning and the University Library. The AI Maturity Scan also helps assess current AI use and provides practical tips for development.
Consult the new framework new framework Generative AI in Education (April 2025) for all principles and guidelines.