England's university regulator called for far greater use of AI-free assessment.

The Office for Students published "Making AI work for students: reflections on institutional practice" at officeforstudents.org.uk on November 11, 2026, which we read there. The page records the research as drawing on conversations with over 300 students and staff from over 120 institutions, focused primarily on English higher education. Press Association coverage, updated at 10:57 Universal Time the same day, gives the figure as 127 universities and colleges; we use the regulator's own wording where we state a number. On assessment, the report says AI poses "significant risks" to assessment validity, that where assessments are unsupervised it is "frequently possible" to complete them with AI without demonstrating the intended knowledge and skills, and that students are often unclear about what AI use is permitted. It says AI detection tools do not consistently and reliably identify AI use, and that "Secure approaches designed to prevent students from using AI will often be the most effective." It records that some institutions have moved to more closed-book examinations, presentations, vivas and other controlled assessments. On study, it distinguishes critical use, such as explaining concepts and generating practice questions, from what it calls cognitive offloading, which it says can undermine learning. On capability, it finds provision of student AI literacy training uneven, staff training often optional with low take-up, adoption driven by individual staff rather than senior leadership, and most staff reporting no significant efficiency gain yet. On regulation, it records that participants generally felt new conditions of registration on AI could be counterproductive and constrain innovation, and that some asked for clarity on complying with existing conditions. Its stated expectation is that "Institutions should develop accessible and robust approaches to assessment, which effectively assure student achievement." Press Association quotes Josh Fleming, the regulator's director of strategy and delivery, saying the position "presents a generational risk to the system as it stands" and that "We need to see far greater efforts to tackle these risks, including far greater use of AI-free assessment," and reports the regulator warning that failure here could breach conditions of registration and lead to sanctions. Not published: any new condition, consultation, deadline, investigation, named institution, or figure for how many students use AI in assessed work.

1. Nov 4

   America's bank regulator told examiners to judge AI by consequences rather than sophistication: the same move from describing a technology to naming the harm it does. Edition 077.

2. Nov 5

   The United States Education Department proposed striking artificial intelligence from its grant priorities, moving in the opposite direction to England's regulator. Edition 075.

3. Nov 11

   The Office for Students publishes its report: detection is unreliable, unsupervised assessment can frequently be completed by AI, and secure approaches designed to prevent AI use will often be most effective.