A new study from The University of Manchester contends that universities are too focused on policing AI misuse and not focused enough on preparing graduates for a working world in which AI is ubiquitous, arguing that the more urgent challenge is equipping students with capabilities that AI cannot replicate.
The paper, authored by Dr Kelechi Ekuma of Manchester’s Global Development Institute and published in the journal Frontiers in Education, argues that employability in an AI-driven economy will depend less on technical expertise and more on critical thinking, ethical judgement, communication, and the ability to navigate complex, ambiguous situations. Graduates entering careers in government, public services, international organisations, and consultancy will encounter AI-powered systems regardless of their academic background, making AI literacy a cross-disciplinary necessity rather than a specialist concern.
Ekuma argues that universities have leaned too heavily on plagiarism detection tools and coursework policing as their primary response to AI, when the more meaningful intervention is redesigning how students are assessed altogether. Oral examinations, reflective accounts of AI use, collaborative projects, and real-world problem-solving exercises are proposed as better measures of distinctly human capability.
The study also calls for AI to be integrated into degree courses across all disciplines, not confined to technology programmes, so that students develop the critical distance to question AI-generated information, identify its limitations, and apply independent judgement where it matters most.
