Whose Work Is It?

Students will meet the AI detection game twice: once on an assignment, and again on a first job application. The line that matters is not AI or no AI. It is whose work it is.

A staffing professional who works with both executives and hiring managers every week told me something recently. AI is coming up more and more on both sides of the table. Companies are using it more to screen applicants and run interviews. Candidates are using it to write their resumes, submit applications and prepare for interview questions. Employers learn to spot the AI submissions. Candidates learn how they are being caught and find new ways around it. He called it a cat-and-mouse game.

If that sounds familiar, it should. It is the same game a lot of classrooms are playing, and your student is going to meet it twice: once on an assignment, and again on their first job application.

Where the game starts

For many students, the first job application comes during high school. Finding options, researching an employer, filling out the application and practicing for the interview all take guidance. Writing a strong resume is a skill worth learning, and many students are now learning it with AI open beside them.

The habits they build there are the ones they bring to the table. So it is worth knowing what the people on the other side of that table are actually asking.

What the other side is asking

In a 2026 Resume Genius report, 58% of hiring managers said they had encountered AI-generated resumes or cover letters, and 46% had seen candidates use AI to answer interview questions. Eighty-seven percent use AI themselves in at least one part of hiring.

So employers are not asking whether AI was anywhere near the application. Most of them use it. In a 2025 TopResume survey of 600 U.S. hiring managers, 52% said using AI for proofreading or drafting support is acceptable, and the report adds that employers expect the final product to sound human, show real effort and reflect the person behind it. About one in five said they would reject an AI-generated resume or cover letter outright.

Read that way, most of them are not drawing a line between AI and no AI. They are drawing a line around whose work it is.

That is the line a school draws too. Our Issue 9 put it this way: an answer is not understanding.

The question that detection answers

Detection asks one thing: did a machine touch this?

That is rarely what a teacher or an employer needs to know. What they need to know is whether this person understands the work, and whether what they are saying is true. A detector cannot answer either one. Worse, it becomes the test. Every tell it finds teaches students what to hide, and soon the work is about passing the detector instead of showing they can do the job. Schools have seen this loop before. When standardized test scores became the goal, many classrooms started teaching to the test, and teachers I know say students learned less and thought less because of it. A student who learns that the goal is not getting caught learns to study the detector, not the subject matter.

The game also leaves honest students guessing. If nobody has made it clear where the line is, a student who uses a tool to fix one paragraph and a student who has it write the whole paper can look the same on the page. One did the thinking and got help with the wording. The other handed off the learning entirely. Without a written line, the honest student has no way to show the difference, and the other one may never be asked to.

This is not a hypothetical. Posts circulating online right now promise students that AI can write their entire college thesis in a few days, brag that the advisor never noticed, and offer the prompt to copy. Some do not even come from real students, just marketing dressed up to look like one. That does not change what they teach. They are being shared with high school and middle school students as a picture of what winning looks like.

The line exists. It just is not written down.

I looked for data on how many employers tell applicants, in writing, what AI use they accept. The only figure I found comes from one region of England, where fewer than one employer in ten had published guidance for applicants. I could not find a U.S. number. If the rule is not written where a student or applicant can read it, then they are forced to guess.

Follow the Rules, our fourth principle, says rules at school and at work come first, and that when they are unclear, a student should ask before using AI. That only works if someone can answer the obvious question, "What are the rules?"

What is in your hands

If you teach, write the rules before you reach for a detector. Three short statements, as a classroom standard or at the top of an assignment, are enough: what is fine, what you want disclosed, and what is off limits. Then put the weight on something a tool cannot hand over. Ask the student to walk you through a choice, or defend it when you change one fact.

If you are a parent, sit with your student for their first application. Let them use the tool if the employer allows it. Then give them one test: could you defend every sentence on this page in your own words, without the tool open? If yes, it is theirs. If not, they are not ready. One follow-up question in the interview is all it takes to find out, and that is a hard way to learn it.

Then practice the interview at the kitchen table. Start by asking why they chose that job, and make sure the answer is their own, not a line they memorized. After the practice, ask what they would do differently, what they learned and how it felt. A student who can only recite a resume is not ready for the room. A student who can explain one is.

The question to take back

If your student asked you what the rules are for using AI on their next assignment, or on their first job application, what would you say?

If the answer is a long pause, that is not a student problem. It is a missing standard, and it is probably missing in more places than one.

AI use is changing outside of education too, and hiring shows how fast. The learning curve is steep for everyone, employers included. Parents and educators can help students find their own line, one opportunity at a time. The machine does not sign the application. Your student does.

What the checking turned up

Six things about the evidence behind this piece, because you should be able to go look.

The cat-and-mouse account comes from one staffing professional describing what he sees. It is an observation, not a statistic.

The point about teaching to the test comes from teachers I know describing what they saw. It is an observation, not a study.

The posts promising students a finished college thesis are ones I have seen circulating on social media. Their claims cannot be verified, and some are labeled as AI-generated. It is an observation, not a measurement.

Both surveys come from companies that sell resume services. Both were run as online Pollfish panels, and neither publishes a margin of error. The Resume Genius report gives two different sample sizes for the same survey, 1,500 in its introduction and 1,000 in its methodology, and does not explain the difference. TopResume reports its rejection figure two ways: rejecting an "AI-generated" resume or cover letter, and rejecting an application that "appears to be fully AI-generated."

Resume Genius, 2026 AI Impact on Hiring Report: resumegenius.com/blog/job-hunting/ai-impact-on-hiring-2026

TopResume, Where Employers Draw the Line on the Use of AI in Hiring: topresume.com/career-advice/ai-in-hiring-survey

These surveys measure what hiring managers say they have seen or would accept. Neither checks whether the material they saw was in fact written by AI.

The figure on published guidance for applicants comes from a University of York study of employers in one region of England, and it does not state how many employers responded.

University of York, What employers really think about AI in job applications: luminate.prospects.ac.uk/what-employers-really-think-about-ai-in-job-applications

When we get something wrong in a published issue, it will be listed here too.