If AI Can Do the Assignment, What Exactly Is University Assessing?

AI can now write essays, generate code, solve problems and help produce polished university assignments. So an important question is emerging: if AI can produce the final answer, what exactly are universities assessing? In this video, I discuss why generative AI is not simply a cheating or plagiarism problem. It raises a deeper question about assessment design, learning and graduate capability. I look at recent evidence on student AI use and examples of how universities are beginning to rethink assessment, before proposing three capabilities that I believe future assessment should increasingly examine: Can the student do it? Can they use AI to do it better? Can they recognise when AI is wrong? The challenge is not to educate students for a world without AI. It is to develop graduates who can think independently, use AI intelligently and exercise professional judgement. Topics covered include: AI in university education, generative AI and assessment, student learning, critical thinking, AI literacy, graduate employability and the future of higher education. Sources discussed: HEPI Student Generative AI Survey 2026 Stanford University work on assessment in the AI era UCL guidance on generative AI and assessment I’ll be exploring more questions around AI, university education and careers on this channel. If you have experience using AI as a student or educator, I’d be interested to hear your view in the comments. #ArtificialIntelligence #HigherEducation #GenerativeAI