The Office for Students warns that artificial intelligence makes fair grading of student coursework nearly impossible. England’s higher education regulator states that students using tools like ChatGPT often achieve similar grades to those working without assistance. A recent joint survey of over 300 respondents across 127 institutions confirmed that AI usage remains undetected in many essays and dissertations. Some students wrote entire papers with generative models and could not recall basic content afterward. Instructors also noted large batches of uniform submissions sharing identical sources and style. Strategy director Josh Fleming described this as a generational risk threatening the credibility of degrees. The agency demands far greater use of AI-free assessment methods. The OfS now threatens to sanction institutions that fail to develop reliable exam formats. Britain’s Department for Education expects universities to take this issue seriously. The regulator wants its framework to be explicit, strong and durable regarding AI use.
This shift forces universities to reconsider how they validate learning outcomes in an automated age. Traditional written assessments are becoming vulnerable to external generation without clear detection mechanisms. Institutions must pivot toward oral exams or process-based tasks to ensure students retain knowledge.
- OfS strategy director Josh Fleming flagged the issue as a generational risk
- Survey data shows undetected AI use in essays and dissertations
- Regulators threaten sanctions against institutions failing to adopt new formats
