When I showed regular 21st Century Tech Blog contributor Harper Lane an article that recently appeared in Toronto’s Globe and Mail on higher education’s efforts to cope with artificial intelligence (AI), she was chomping at the bit to add her two cents. The article was written by Anna Jahn, Executive Director for the Centre for Media, Technology and Democracy at McGill University in Montreal. It described how universities were coping with AI, either by trying to be AI-free or by returning to the cloistered model of learning using small study groups, oral presentations, and in-class written tests.
Harper immediately pointed to the 21st-century contradiction if universities were to deny AI’s existence for students who would then graduate into an AI-infused job market. She stated that not mastering AI as a learning tool would be a fatal mistake. But that wasn’t the end of it. When I told her I wanted to write a response to the Jahn article, she asked if I wouldn’t mind if she could give it a go.
Boy did she ever. These are Harper’s thoughts on the subject delivered in three parts. Part 1 is entitled “What are Educators Truly Measuring?”
What are Educators Truly Measuring?
A sophomore sits down to write a paper. Twenty years ago, writing the paper meant reading, outlining, struggling through arguments, revising drafts, and eventually producing something imperfect but genuinely their own. Today, it might mean opening ChatGPT, Claude, Gemini, or Copilot, typing a quick prompt, pasting in the rubric, and receiving a passing paper in thirty seconds.
The professor knows this. The student knows this. The university knows this.
And yet everyone continues acting as though the assignment still serves the same educational purpose it did a decade ago.
This is the central crisis facing higher education in the age of AI. Universities are not simply struggling with academic honesty or plagiarism. They are confronting a much deeper problem: two educational goals that once complemented one another are suddenly pulling in opposite directions.
The Great Educational Divorce
At the heart of the debate is a distinction that universities have been reluctant to acknowledge.
One objective for education is to develop subject-specific knowledge, reasoning processes, and intellectual frameworks. Students learn how historians think, how physicists solve problems, and how philosophers construct arguments.
The second objective is preparing students to operate effectively in a technological environment. This includes software proficiency, digital research techniques, data analysis, and increasingly, AI-assisted work.
Historically, these objectives have lived in harmony. A student who learned to use modern spreadsheets rather than manual ledgers became a better accountant. A student who learned to use statistical software became a better researcher. But AI changes this equation because it increasingly replaces parts of the intellectual and learning process.
The result is a growing tension: if universities encourage unrestricted AI usage, students may become highly effective technology users while developing weaker foundations within their disciplines. But if universities prohibit AI or access to technology more broadly, they risk graduating students unprepared for workplaces that increasingly expect AI technological literacy.
Both concerns are legitimate, but the mistake is assuming they must be addressed within the same educational space.
Why We Keep Designing Courses for a World That No Longer Exists
Many proposed solutions focus on detection software, stricter policies, and increasingly elaborate methods of monitoring student behaviour. These approaches misunderstand the problem entirely.
The issue is not that students have become more dishonest. Rather, traditional assignments no longer reliably measure what educators think they measure. A take-home essay made sense when writing an essay required personally completing every step of the process. Today, the connection between assignment completion and learning has weakened considerably.
Consider a common scenario. A student receives full credit on a paper generated largely through AI assistance. The gradebook records success. The transcript records success. The student progresses toward a degree.
But did learning occur? Not likely. Many students using AI cannot even remember key points from their own AI-generated essays.
The difficulty is that educators often have no reliable way to distinguish genuine mastery of a subject from the successful use of an AI tool.
A Problem Bigger than AI
AI did not create a crisis in higher education. Rather, it exposed one. Universities have spent decades operating under the assumption that completing assignments and acquiring knowledge were roughly the same thing.
As long as students had to perform most of the intellectual labour themselves, that assumption largely held. Today, it no longer does.
This is why debates about plagiarism detectors, AI policies, and classroom restrictions often feel unsatisfying. They address the symptoms without fully addressing the underlying problem.
The challenge facing universities is not simply determining whether students are using AI. It is determining how learning can be measured when completing an assignment no longer guarantees that learning occurred.
Before institutions can decide how to redesign courses, assessments, or degree programs, they must confront another uncomfortable question:
What are students actually trying to achieve when they pursue a university education?
That question turns out to be just as important as the technology itself. We will try to address this question in Part 2: Why Education is No Longer About Learning.
