Saturday, 27 December 2025

External examiners reports and the GenAI issue

I was grazing through one of the many mathematics forums a few days ago and I stumbled on a post that suggested that the external examiners’ reports for the last academic year (24-25) have now been released, so I had a look for myself. Sure enough, they were there for most curriculum areas, so I delved into the mathematics and physical science ones, since these are applicable to my course of study. The common theme in all the reports that I saw appears to be that GenAI is a massive problem for examinations taken remotely, and the recommendation being that there should be a return to face-to-face invigilated exams as soon as possible to avoid OU degrees being regarded as sub-standard. It was also apparent that the OU senior management were very reluctant to do this, almost certainly due to cost and logistical reasons. I honestly think they don't have enough staff to make the arrangements, as the exams used to be organised on a regional basis and the regions were dissolved a few years ago. I don't think any of the remaining centralised staff have the capacity to do it.

Of course, it's relatively straightforward to check a wordy essay for plagiarism or AI assistance by running it through the appropriate software, such as Turnitin, but science and mathematics-based content is much trickier to scrutinise since the answers tend to be either correct or incorrect, with very limited ways of reaching the same result or conclusion. I'm alarmed therefore at the number of students who have been subjected to an academic conduct review on the grounds that their submitted work has been plagiarised. From looking on social media, it looks like the OU have been somewhat overzealous on occasions in an attempt to maintain academic standards and shore up assessment integrity. According to some Reddit and Discord punters, the 'evidence' that the OU have used to make the allegation of cheating is flimsy to say the least, and tends to be something along the lines of 'we ran your answer through our magic software that detects plagiarism and you used a phrase that wasn't in the textbook'. No hard evidence is presented to the student, and the accusation vague enough to allow various interpretations about the specific wrongdoing.

It reminds me a bit like the old days when disinformation was disseminated that 'TV detector vans are operating in your area' and could tell if you were watching the telly without a licence. The truth was that the GPO (later outsourced to Crapita) had a list of all the postal addresses in Britain and another list of all the addresses that had a TV licence. They subtracted one from the other and banged on a few of these non-licenced doors hoping to catch the occupants with Coronation Street on in the background. The vans claiming to contain the 'sophisticated electronic equipment' that could detect you watching Mavis getting a touch of the vapours in the Rovers Return, were actually just a few empty metal boxes with analogue dials and flashing lights which looked plausible in the material used for propaganda purposes.

I suspect that many of the so-called academic conduct reviews are similar fishing expeditions, hoping that a few students will be intimidated into 'fessing up' in exchange for getting off with a semi-formal warning but still being allowed to continue with their qualification. The alternative would be to deny the allegation, which would put the release of the module result ‘on hold’ while a thorough investigation takes place. That could take months to complete, and nobody would want something like that hanging over them. There is, therefore, a big incentive to admit to something that may or may not have happened just to get the matter over and done with, even though the validity of the evidence is questionable. I'm not sure I like the sound of that.