Generative Artificial Intelligence and the Risks to Assessment – A Quick Experiment

Interest in the current trend of AI-based everything has surged in recent weeks due to media coverage of various chat bots that have existed for quite some time.  Below is an article that I ‘wrote’ in October 2022 using an online ‘AI’ text generator app whilst I was travelling on a train in Scotland.

Whilst it is by no means flawless, it really is quite clever.  At the time I was trying to assess whether this type of technology would represent a threat to the integrity of the sorts of exams that young people go through in their formative years at school – so GCSE and A-level type assessments.  My opinion is that what this machine churned out in 1 second, is most likely a lot better than the sorts of incoherent waffle that I used to come up with when I was in that age bracket of 16 – 18 and therefore yes, this is a real challenge for anyone working in assessment and awarding.

However, the discussion in online forums has very quickly escalated from the ‘miracle’ of these chat generators, to the existential threat of General Artificial Intelligence and how this may well herald the end of the human species.  Some might suggest that this is due to the culture of mass-anxiety that we currently live in, and they would be correct…  So, bringing our focus back down to Earth and to the point at hand whilst trying to control our fears so that we can mount an effective counter to the emergence of a new risk to assessment integrity; is a good place to start in the defence of humanity against our evil robot overlords of the future.

Of course, the implications of use of generative AI’s to commit academic fraud in exam situations and beyond, are extremely serious.  Discovering that an employee who has been selected for a position – based in part on the qualifications listed on their CV, has clearly lied and cheated their way to gaining the qualification, is not only quite shocking, but it encourages employers to ignore other applicants’ qualifications – regardless of how difficult they were to earn.  Added to this is the fact that there is no mechanism, formal or informal, with which to highlight this type of fraudulent behaviour – meaning that the person is free to go on to apply to any other organisation for any other role, with no fear of exposure of their deceit – and there we have a problem that will not be resolved.

So what is the solution to qualifications that can be earned using AI chatbots, cheat-for-hire companies and other nefarious means?  Humans.

It remains the case, that until we have a true General AI that is capable of understanding the full nuance of human consciousness, responding with empathy and compassion when needed, yet remaining disciplined in the face of potential coercion and deception, then our best line of defence remains ourselves.

People are the answer.  Good people, honest people, people who understand the value of education and qualifications and of the negative effects created by those who wish to cheat their way to the top.  People who have the courage to call out malpractice and fraudulent behaviour for what it is, whilst avoiding hysteria and misplaced zeal.  The answer is not more ‘Ais’ – for example plagiarism checking algorithms that do nothing more than provide a tick in a box for a regulator.  By allowing cheating to continue through implementing ineffective countermeasures, we allow the continuation of the undermining of the value of the assets that we should regard as most valuable – our qualifications.

So we must continue to effectively, independently and rigorously defend the integrity of assessments through high-quality invigilation.  Invigilators are not merely squeaky shoes on the exam hall floor, they are the frontline protection for honest candidates against those who wish to cheat them of their success.

 

GPT Generated Article ‘written’ 14 October 2022

My input:  “Recent years have seen an increase in the means, opportunity and motivation for exam candidates to cheat in order to prosper.  Technology continues to advance at a pace that administrative communities struggle to keep up with as they also try to maintain their day jobs of tackling the exam cheating that they already know about.  But what drives people to cheat in their exams?”

AI output:  “The majority of people cheating in exams are students. They have time, motivation and access to better resources than their teachers.

There’s good evidence that students believe that cheating is the best way to increase their chances of success. They do so because it benefits them in terms of their examination performance, but also because they enjoy the experience of taking part in a morally questionable activity.

Widespread cheating has seen a decline in the quality of Australia’s educational system over recent decades. This decline is in some part due to increased efforts by governments and education authorities to deal with cheating, but it’s also partly due to an unintended consequence of more intelligent students wanting to ensure their success.

A successful exam attempt is a perfect time to cheat

It’s well known that many successful students leave exams early when they know the answers. However, it’s also true that these students leave early if they don’t know the answers. It’s therefore not unreasonable to suspect that many of the successful students leave early because they’ve already decided what the correct answers will be.

One solution is to extend the number of subjects taken at different age levels.

A second solution is to have exams on mobile phones or tablets – a feat that seems impossible due to the security features of these devices.

But it’s also true that some groups of students have the knowledge, access and motivation to cheat – and that those groups are not simply students. It’s worth considering that a reasonable guess is that exam cheating is on the increase partly because it’s become easier and cheaper for students to cheat.

Self-centred motivation

It’s difficult to know for certain whether students are motivated by short-term gain or long-term gain, because these motivations are not necessarily connected to whether students believe the cheating is unethical.

Nevertheless, the evidence suggests that self-centred motives are the main motivation for cheating. Students take an…”

To find out more about our work and how VICTVS is countering the use of AI in exams and assessment, please contact us here or call us on +44 (0) 113 539 7056.

1 Comment. Leave new

Leave a Reply

Your email address will not be published. Required fields are marked *

Fill out this field
Fill out this field
Please enter a valid email address.
You need to agree with the terms to proceed

error: Content is protected