Our approach, and why we've taken it.
The University of Westminster's commitment to the UN Sustainable Development Goals requires us to engage thoughtfully with all technologies that are reshaping society. Generative AI offers real opportunities to improve access to education, enhance accessibility, advance research, and help tackle complex global challenges. It also brings significant challenges, including environmental impacts, bias, privacy, intellectual property, and the potential for misuse.
Rather than ignoring transformative technologies like generative AI, we believe universities have a responsibility to help people understand, question and use them wisely.
Accordingly, we have not banned the use of AI in teaching and learning nor have we left its use entirely unregulated. Instead, we have developed a framework that applies to every member of our University community, students and colleagues alike. We have developed, through consultation of our university community, six guiding principles for the use of generative AI by students and colleagues alike.
Through these principles for the ethical, safe and responsible use of generative AI, we aim to prepare our community to realise its benefits, address its risks and contribute positively to a more equitable and sustainable future.
We know that this matters beyond the university. Parents want to be confident that a Westminster degree reflects genuine capability. Employers want graduates who understand AI, not just graduates who have used it. Our aim is to equip every student with skills and judgement that hold up in the world, and these principles are how we go about that.
Our six guiding principles
1. Use AI Responsibly and Critically
AI should be used where it serves a genuine educational, professional, or operational purpose, not to avoid demonstrating your own learning, judgement, or expertise. Its use should always involve critical engagement, including understanding what the tool is doing, why you are using it, and its limitations.
Why this matters: The question at the heart of this principle is not whether someone used AI, but whether they used it thoughtfully. AI can be an excellent tool for exploring ideas, getting feedback on a draft, or understanding a difficult concept more deeply. It becomes a problem when it substitutes for thinking rather than supporting it. We want our students to leave Westminster not just knowing how to use AI tools, but knowing when to use them, when not to, and how to evaluate what those tools produce. That's skill employers consistently tell us they value.
2. Be Honest and Transparent About AI Use
All members of the University community should be open about when and how AI has contributed to their work, teaching, or decisions, including what was used and how. Concealing substantive AI involvement in the development of an output constitutes misrepresentation.
Why this matters: Trust is the foundation of academic life and of professional life beyond it. If a piece of work has been substantially shaped by AI and that isn't declared, we lose the ability to assess it accurately, to support the student effectively, or to stand behind its integrity. Transparency isn't just a rule; it's how we protect the value of the work everyone produces. Our guidance makes clear when and how AI use should be declared, and we hold all members of the University to that standard equally.
3. Remain Responsible for Your Work and Your Judgements
AI does not take responsibility. Whatever AI contributes to a piece of work, a decision, or a teaching activity, the human who uses it remains fully accountable for the outcome. AI output must always be critically evaluated, not simply accepted.
Why this matters: AI tools can be confident and fluent while being factually wrong, poorly reasoned, or inappropriate for the context. Students whose names are on a piece of work are responsible for it and that means checking what AI has produced, not just submitting it. For colleagues, this principle is equally important: AI can support professional judgement but cannot replace it. Significant decisions, especially those involving student welfare, academic misconduct, or assessment outcomes, must remain human decisions, defensible by the person who makes them.
4. Act with Fairness and Consider Broader Impact
AI use should not create unfair conditions, including those arising from unequal access to paid and free AI tools and from differing levels of confidence and experience among students and colleagues. The University and its community should also consider the broader social, economic, environmental, data privacy, and intellectual property impacts of their AI choices. The university will keep the baseline access it provides under review.
Why this matters: It would be easy to assume that everyone has equal access to AI tools, or equal confidence in using them. That assumption is likely to be wrong, and basing decisions on it risks embedding new forms of disadvantage. At an individual level, we ask everyone to be aware of this. At an institutional level, it shapes how we select and procure AI tools, weighing environmental and social impacts, not just cost and functionality. Suppliers' commitments in these areas should be genuine and verifiable, not simply aspirational statements.
5. Support Institutional Learning and Adaptability
The University’s approach to AI should be subject to regular, at least annual review, informed by evidence from practice, and open to revision. This applies to guidance and support for all colleagues, academic, professional services and other technical colleagues as well as students. Everyone should have meaningful input into the evolution of policy and guidance, including through structured training and development.
Why this matters: AI is developing faster than any policy or guidance document can keep up with. We do not pretend to have all the answers, and we recognise that the people closest to practice, students working with AI in their studies and assessments, colleagues integrating it into their teaching, will see things that policy-makers do not. We are committed to creating genuine channels for that experience to shape our approach, not just consulting for the sake of it, but listening and responding. Good policy in this area must be built on what actually works, not just on what sounds reasonable in theory.
6. Protect and Develop Human Capabilities — Including Creative and Intellectual Originality
AI use in education should not erode the skills, knowledge, and capabilities that higher education exists to cultivate. This includes critical and analytical thinking, creative and original thought, the ability to evaluate evidence and argument, and the confidence to keep developing throughout a career. In some disciplines, the act of making, interpreting, or arguing is itself the major capability being developed; in those contexts, the question of what AI displaces requires particular care.
Why this matters: This is perhaps the most important principle of all, and the one that shapes everything else. A university degree is designed to develop capabilities that serve a person over a lifetime, not just to produce outputs at graduation. If AI does the thinking, making, or arguing that a course is designed to develop, the student gains a qualification without gaining the capability it is meant to certify. This question has additional weight in disciplines where a student's own voice, interpretation, or creative vision is the substance of the work. For example, in the creative arts, media, architecture, humanities, and social sciences. Our assessment design across all disciplines aims to keep the student's own contribution visible, and to ensure that where AI is used, it genuinely supports rather than replaces the capabilities we are here to develop.
What this means in practice
If you're a student
Before using AI on any piece of assessed work, check the assessment brief to see whether it is permitted, restricted, or prohibited. If you use AI in a way that meaningfully contributes to your work, declare it as set out in the University's guidance. Never enter personal data about yourself or others into external AI tools. And ask yourself honestly - does my use of AI deepen my learning, or is it doing my thinking for me?
If you're a parent or guardian
These principles are designed to ensure that your children’s degree reflects their own genuine capabilities and that they leave Westminster equipped with the kind of critical, reflective approach to AI that employers increasingly look for. The framework doesn't treat AI as the enemy, it treats it as a tool that needs to be used well.
If you're an employer
Westminster graduates are taught to engage with AI critically and responsibly, to use it purposefully, to take accountability for the work they produce, and to understand its limitations. That is the kind of AI literacy that we believe adds lasting value in professional contexts.
If you have questions about how GenAI is used at Westminster, please email [email protected].