England's university regulator says AI makes grading coursework "close to impossible" and recommends a return to "AI-free assessment" methods like written exams
First reported by The Guardian ·
The ability to use AI tools to complete university coursework is no longer a viable path to a good grade.
England's university regulator, the Office for Students (OfS), has declared that artificial intelligence has rendered coursework assessment "close to impossible." In a report based on interviews with over 300 students and staff across 127 institutions, the OfS found widespread AI use by students to complete essays and dissertations. This has raised concerns about the validity and reliability of degrees, with some academics reporting that submissions are increasingly indistinguishable and difficult to assess for genuine student knowledge. The regulator is urging universities to revert to "AI-free assessment" methods, such as traditional written exams and in-person presentations, to safeguard academic integrity and the credibility of higher education. Failure to adapt assessment strategies could result in sanctions for institutions.
The OfS's directive signals a significant challenge for higher education institutions grappling with the rapid integration of AI. While some universities have already begun modifying their assessment methods to mitigate AI misuse, the regulator's strong recommendation for a return to traditional, AI-free formats implies a potential industry-wide shift. This could necessitate substantial investment in new examination infrastructure and a re-evaluation of pedagogical approaches to ensure that learning outcomes can be authentically measured in an AI-influenced world.
This situation creates a complex landscape for both students and educators. Students who relied on AI for assignments now face a future where such methods may be ineffective or actively penalized, potentially requiring a renewed focus on original thought and critical analysis. For faculty, it means developing and implementing robust assessment strategies that are demonstrably resistant to AI manipulation, while also exploring how AI might be ethically integrated into the learning process rather than solely as a tool for cheating.
AI-written summary. May contain errors.