UK regulator threatens universities with sanctions unless they adopt AI-proof exam formats

England's higher-education regulator, the Office for Students (OfS), has warned universities that they could face sanctions unless they develop reliable assessment formats that resist AI misuse, according to reporting by the Guardian. The regulator's strategy director, Josh Fleming, described the situation as a "generational risk" to the credibility of degrees and called for "far greater use of AI-free assessment." The OfS position is that AI has made it "close to impossible" to grade student coursework fairly.
The warning rests on a joint survey by the OfS and Advance HE covering more than 300 respondents across 127 institutions. It found heavy student use of AI for essays and dissertations, frequently going undetected, with grades comparable to those of students who worked without it. The report describes cases where students produced entire papers with ChatGPT and afterward could not recall basic content from them. Instructors also reported batches of uniform submissions that shared identical sources and style.
Those findings point to a specific problem with take-home written work: it is hard to verify authorship, and the output can be polished enough to score well even when the student has not absorbed the material. The regulator's answer is to push assessment back toward formats where the work happens under observation or leaves a visible trail — traditional exams, in-person presentations, oral exams, process-based assessments, and tasks where students critique or revise AI-generated text. The article notes that many universities are already shifting in these directions.
The Department for Education said it expects universities "to take this issue seriously" and wants the OfS to ensure its regulatory framework is "explicit, strong and durable on the use of AI." The OfS has said it could sanction institutions that fail to develop dependable formats.
For product and platform teams in education technology, this signals a likely demand shift. Tools built around take-home essay submission and similarity checking address only part of the problem described; assessment formats grounded in supervised or process-visible work — recorded oral exams, in-person presentations, versioned drafting, and revision exercises — may see more institutional investment. This is my own reading rather than a stated OfS position, but it follows from the formats the regulator is endorsing and from the survey's finding that undetected AI use can earn ordinary grades.
Why it matters: universities in England face regulatory pressure to change how they assess students, and vendors serving them will need to support exams, presentations and process-based work rather than essay pipelines alone. For students, the practical effect is likely to be more supervised and oral assessment. The exact sanctions and timelines remain open, since the reporting describes a warning rather than a published enforcement schedule.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
England's higher education regulator, the Office for Students, says AI has made it "nearly impossible" to fairly assess student coursework. An investigation found that students are having ChatGPT write entire papers undetected, without being able to recall basic content afterward. To save the credibility of degrees, universities now have to return to traditional exams or face sanctions. The article UK regulator threatens universities with sanctions unless they adopt AI-proof exam formats appeared first on The Decod