Don’t worry, this post isn’t about Artificial Intelligence. No, this is about that other, far more damaging A.I. - Academic Integrity.
Take, for example, these two sentences:
It was hard to believe the impossibilities that lay before her.
It was hard to believe the impossibleness that lay before her.1

With standardized English tests like the TOEIC and EIKEN, only one of those is correct. And yet, you understood them both, didn’t you? You might feel one is more poetic or one is more interesting. But, according to the tests, only the first is correct2.
This creates a lot of pressure for language learners. And, to be fair, a lot of recent neurological research has shown that some pressure is good. Truthfully, the core of classroom management is finding the sweet spot between safe space and pressure cooker. Too much of the former and nothing gets done. Too much of the later and chaos reigns. So, while pressure can be a good thing, too much of it causes learners to look for release valves. Often, the easiest, most direct release comes through some form of cheating.
And that’s important because there is still a popular idea that the students who cheat are those who were too lazy to study properly, just as there is a popular idea that athletes cheat because they care about winning more than playing fair. In both cases, of course there are outliers for whom those are exactly the reasons they cheat, but for many, many more people, they cheat because the pressure to succeed is overwhelming.
My university, like every other institution on the planet, is grappling with the realities of A.I. How much use of artificial intelligence is acceptable? Which uses? In what situations? And, most importantly, when is the use of A.I. a violation of academic integrity?
Let’s begin by taking a look at the fundamental word underlying our case, integrity. Merriam-Webster gives us three separate senses of integrity:
a) firm adherence to a code of especially moral or artistic values
b) an unimpaired condition
c) the quality or state of being complete or undivided
The divide between these three definitions is stark when we look at where the word integrity comes from. Originally from Latin, where it meant whole or complete with a metaphorical connotation that leaned into purity and correctness, it entered English in the 13th century, but by just a hundred or so years later, it had already begun to broaden into something like today’s usages. Per Etymonline:
The sense of "wholeness, perfect condition" is attested from mid-15c.; that of "soundness of moral principle and character; entire uprightness or fidelity, especially in regard to truth and fair dealing" is by 1540s.
These differing meanings alone can make it hard for students to get a handle on what academic integrity means. After all, when we look through the corpora3, we can see that the most common collocates are all over the place. Starting the list with more than 35,000 instances across over a billion indexed words is the collocation territorial integrity while a phrase oxymoron like congressional integrity4 is waaaaaay down near the bottom with only 124 occurrences across that same billion words. Scattered throughout the middle we have phrases that match all three meanings of integrity.
Starting with definition 3 (complete or undivided) and with no attempt at any sort of order, we have: complete, overall, total, institutional, organizational, and, of course, territorial.
Then, moving up to definition 2 (unimpaired), we find words like, structural, physical, mechanical, material, mental, and even historic.
But, obviously, it’s definition 1 (moral correctness) that we’re most interested in. Here we have phrases like, well, moral, ethical, spiritual, professional, and, of course, academic.
So, for students, a poorly written academic integrity document can be a minefield. Are they meant to read it as a do/don’t-do list of rules? Is it a guide for being a “complete” student? Does violating academic integrity mean becoming somehow less than whole5? While I’m obviously exaggerating the linguistic confusion, the chances for misunderstanding what constitutes morally correct behavior is real.
Especially when it comes to brand new technologies that are rapidly becoming baked into every bit of personal and classroom tech the modern student needs. So when the Academic Integrity guidelines state “No A.I. use is permitted,” how are the students to know where the A.I. line actually resides? Is Google Translate powered by A.I.? (Yes.) Is the spell check in Microsoft Word? (Depends on which version of Word you’re using.) Is the grammar checking software you’re using to double-check that your gerunds are gerunding appropriately powered by A.I.? (Probably.). You get the idea.
So where does this leave us? It leaves us in a precarious situation where the people making the most about students using artificial intelligence to violate academic integrity are the people who seem to least understand just how pervasive the technology is becoming. They’re writing guidelines with no clear idea of what actually constitutes a violation, making the rules for students and teachers alike a grey area full of uncertainty and confusion. And, frankly, that is less than morally sound.
Stay curious,
Joel
Hint: They’re both right if for no other reason than you can’t prove that they’re not.
Okay, fine. The second one is also correct, grammatically, but impossibleness is an archaic form of impossible that really isn’t used much anymore. Since the standardized tests aim for modern, business-level, international English, they’re mostly justified in creating problems like this.
I used the NOW Corpus of over 16 billion entries via the English-Corpora website for all the data used below.
This is called a prejudicial reading and when it comes to the U.S. Congress, guilty as charged.
While there are entire branches of philosophy that would argue, yes, those branches tend to be populated exclusively by dead guys whose opinions on modern life can be less than well-informed.
I explain my experiments with A.I., what I learned and how Iearned it in my podcast here:
https://open.substack.com/pub/soberchristiangentlemanpodcast/p/ai-deception-2025-pt-2?utm_source=share&utm_medium=android&r=31s3eo