Since the 2026 semester has started, I’ve noticed a disturbing trend: I’ve been receiving regular emails from students that I’m convinced are AI-generated.
This presents a strange conundrum to me. Using AI to write emails is not strictly forbidden at the institutions at which I teach the same way using AI to generate an assignment is, but in many ways, I find this trend to be more concerning than AI assignments, and I'm not a fan of those either.
While I disagree with using AI to cheat on writing assignments, I understand the rationale. Writing assignments are difficult and time-consuming, and students often believe ChatGPT can do a better job than them in a few moments of work.
Writing an email can be annoying, but for most college-level writers, it is not a particular challenge, so having a computer impersonate you and write one is another level of inauthenticity in my mind. This is one that goes beyond merely cheating on a test and feels more fundamentally off on a human level.
Put another way, getting AI emails make me sad and is a little disturbing. But that's a strong reaction to spring on a student for using a tool that is now often integrated with their school-issued email platform, so here's how I am responding to the trend of AI-generated emails.
1. Understand The Perspective Of Students Who Use AI To Write Emails
Much as I hate this latest trend, I know students, in many cases, are being actively encouraged to use AI by tech companies, and human writing is increasingly looked down upon by many outside the education field--and even some within it.
In other words, having AI write your emails for you is a semi-accepted practice in modern society. So calling out individual students and/or others isn’t helpful or fair, and might make things worse.
Instead, I'm trying to ask more questions and learn about the mindset that is leading students to turn to AI-generated emails.
2. Call Out The Content, Not The Origin
That said, while responding to student emails, I think I'm perfectly justified in highlighting strange AI-generated phrases. I may ask a student to offer more specifics or point out exactly why a particular sentence or phrase doesn’t make sense.
Here is an excerpt from an email I generated with AI as an example. In my prompt, I asked to ask my professor for an extension on an upcoming assignment because of all the recent snow in my home state. The response was typical of what I've been seeing lately, and reads in part as follows:
We’ve been experiencing significant snowfall here in Connecticut, and the weather has made it difficult to maintain my usual work schedule and access quiet, reliable time to complete the assignment to the standard I aim for. I want to be sure I submit work that reflects my best effort and understanding of the material.
In my response, I’d ask them to describe what their definition of “reliable time” is, and ask them to tell me a little bit more about the standards they for which they are aiming for, before granting such a request.
3. Model Authentic Behavior
Many books have started to include an authenticity statement in which the human-written nature of the work is asserted. I think this is something those of us who care about genuine human expression should consider doing the same.
It can be as simple as adding an email signature such as: "This email was 100 percent human-generated."
4. Fight The Culture of AI Email Writing
It’s time for those of us concerned about AI writing to get more vocal. AI writing is bad writing, period. It doesn’t enhance human writing or communications skills, and, instead, robs human warmth and voice.
More to the point of this article, using AI to write an email is inauthentic, and I’d even argue fraudulent. I'm all for saving time and cutting back on the annoying emails many of us need to keep up with to function in modern society, but this isn’t any kind of solution.
5. Offer Alternatives To AI Email That Save Time
To avoid spending this whole piece just complaining, I think the problem of AI emails, unlike AI-generated student work, has an easy solution: We should normalize the use of AI assistants who speak for us, not as us.
If a student's email began “Hi, this is your student Wesley Crusher’s AI assistant. Wesley is confused about the recent midterm assignment . . . .” I’d be a little surprised, but I’d have no personal problem with this type of AI use. In fact, I might have my own AI assistant send a response.
I’m all for using AI to save time in otherwise mundane and informative communications, but let’s cut this AI-human impersonation nonsense off before it goes much further.
