Opinion
Cheating with AI: An impending collapse of higher education?
NOVEMBER 22 — Like chess, education faces an existential threat in the form of cheating.Unlike FIDE (the World Chess Federation), though, educational institutions (not least those in Malaysia) are much more lax about the situation.I’m talking about the illegitimate use of AI by students for their assignments, projects, etc.The (by now ubiquitous) use of Gemini, Co-Pilot and similar programs to generate answers and essays which are subsequently submitted as students’ own work renders suspect the validity and integrity of all digitally-produced assessments.According to a recent investigation by the UK’s Office of Students (an independent regulator of higher education in England), AI has made it "close to impossible” to assess university coursework. Which, in turn, threatens to destroy the (already vulnerable) credibility of degrees educational institutions offer.The problem is easily understood yet, speaking as an insider in the education sector, nobody’s really doing anything substantial about it.The usual "resolutions” don’t help much, as these are usually:Invalid: No AI-detection software can definitively prove that a student used AI; unlike plagiarism checks which can pinpoint the exact website and page of something a student has copied, AI-detection software can only give an estimate that a certain paragraph (many of which look completely normal, to be honest) has been generated by DeepSeek or whatever.Impractical: Hand-written exams in closed-door settings may work, sure, but can institutions do this for every assessment in every subject? What about non-exam assignments? Group projects?Outdated: Closely related to problem #2, are we really NOT going to use, uh, computers for written assessments in universities? Are lecturers supposed to revert back to the 80s’ where students submit everything via paper?Limited: The least problematic cases of AI use are when lecturers can perform "manual” evaluations on students (in the form of vivas or presentations and so on); however, and without denying the importance of this approach, no one can deny that this could only feasibly apply to small class populations (I mean, simply try interviewing 40 students for every assignment they do, let alone 400?)Inimical: If an institution brings in thousands of foreign non-English speaking students who use AI-powered (and is there any other kind?) translation software, the game is over from the start; AI-detection software has enough trouble "confirming” that a student copied from Grok let alone deciding that certain sections are merely "translated” from another language.In the final analysis, perhaps hand-written exams are the only way. This is how guys like me did their UK Bachelor degrees in the past: We studied for three years and we "do or die” in one three-hour closed-door exam in a hall big enough for 13 volleyball courts.Then again, those are degrees from selected (and not all) UK universities. What about qualifications from other places? What do we do then?Again, since late 2022 when ChatGPT was first launched, everybody in education surely had to know this was an issue they cannot ignore for long.When will the government, parents and other stakeholders fly the red flag and demand some answers?Should a moratorium on education be put in place until some standardised policies have been hashed out and every institution put on notice?Failing a drastic rethink, the sector’s "sub-prime”-like crash and collapse are nothing if not imminent.* This is the personal opinion of the writer or publication and does not necessarily represent the views of Malay Mail.
You May Also Like