Professors do misinterpret good writing as AI-generated. As a matter of fact, dramatic improvements in a student’s writing—or writing that seems “too good”— are among the most common grounds to suspect that a student used AI (against the professor’s rules).
As a student, you should hear this: Quality writing should not be held against you. Furthermore, an instructor’s suspicion that you used artificial intelligence to write on your behalf is not sufficient proof of misconduct. Yet your school may be able to impose discipline even without concrete evidence that you did something wrong.
As soon as you recognize that you are suspected of illicit AI use, reach out to the LLF National Law Firm Professional License Defense Team to put your defense in place.
If you have not used AI as you’ve been accused of, we will seek to alleviate your professor’s concerns. If you did use AI, perhaps even in violation of your professor’s rules, we will seek to mitigate any harm you face as a consequence.
Misconduct allegations are nothing to take lightly. Call the LLF National Law Firm Student Defense Team at 888-535-3686 or contact us onlineto discuss how we can help you.
Why Might a Professor Misinterpret Good Writing for AI-Generated Content?
Let’s put ourselves in the perspective of a professor who is reading their students’ written submissions. That professor is likely to be:
- Evaluating the student’s writing as professors always have, gauging the accuracy of grammar, sentence structure, and punctuation
- Evaluating how compelling the student’s arguments and points are
- Looking for any indication that the writing is not the student’s own (but instead work created by a third party, including artificial intelligence)
When trying to determine if the writing in question has been created by the student, AI, or some other party, the instructor may consider:
- The general quality of the writing: It is a jaded perspective, but one that professors are not entirely irrational to have. When a student submits a written assignment that stands out in a positive sense, many instructors at least consider that the work could be the product of artificial intelligence. AI-generated writing is not perfect, but it tends to be polished and error-free.
- The student’s past writing: Good writing may be more likely to strike an instructor as abnormal if the student’s past submissions have been…not so good. Writing that is drastically improved tends to command more suspicion than writing that is simply good.
- Repetitious or “predictable” elements: Tools trained to identify AI-generated writing look for several tells. Two of the features that these AI detectors look for are repetition and predictability. An instructor who is familiar with the hallmarks of AI-generated writing might look for the same features. Yet, a student can intentionally or unknowingly engage in repetitive writing—whether because that’s just their style, they are trying to emphasize a point, or they believe it is a sound way to structure their work.
- Particular types of punctuation: Those familiar with AI-generated writing know that certain types of punctuation appear unusually often. Certain types of dashes, for instance, are associated with AI-generated work. Yet, human beings are just as capable of using the same punctuation in their work—just trust me on that.
There is no perfect means of differentiating AI-generated writing from human-crafted work. An instructor’s powers of observation are certainly not enough, and the tools they might use to stress-test their suspicions are similarly imperfect.
Do Instructors Have Any Way to Verify (or Refute) Their Suspicion That a Student Used AI?
Yes, there are ways for instructors to try to verify or disprove their suspicions that a student used AI to write their work. These verification methods are not perfect, though.
The class of tools commonly known as AI detectors is the primary method for detecting AI-generated work. Some examples of this technology are:
- GPTZero
- ZeroGPT
- Turnitin AI Detector
- Copyleaks
These and other products that promise to identify AI-generated writing vary in their accuracy. Not a single one of them can boast 100% accuracy—and many fall far short of perfection.
AI detectors are not only capable of raising false flags and failing to flag AI-generated work. These detectors are also known to be biased towards certain groups, particularly those whose writing may mirror trends seen in AI-generated work.
This is why, outside of a student’s confession that they used AI to write for them, there is no foolproof way to prove as much.
How Can I Prove That I Did Not Use AI to Write the Work in Question?
An instructor or other party at your school does not necessarily have to prove that you committed AI misconduct beyond a reasonable doubt.
Most universities apply the preponderance-of-the-evidence standard in adjudicating alleged AI misconduct. In other words, if the adjudicative body believes it is most likely that a student had AI write their paper, then that student may be disciplined.
For an accused student, this means that a proactive defense is the right approach. Ideally, you would prove that you did not commit the AI-facilitated misconduct you were accused of.
A student might prove that they wrote the work without AI’s help by:
- Verbally describing their research and writing processes
- Volunteering to answer questions from investigators about the paper’s contents
- Providing any digital records, such as proof of the draft being completed portion-by-portion, which proves human creation
- Citing any past work that contains similarities to this work, which could aid the student’s case that this is their writing (not ChatGPT’s)
If a student did use AI to write an assignment, they will need to decide whether they want to admit it. In some cases, it is wise to admit wrongdoing and seek a merciful disciplinary decision (or possibly even a warning). In other cases, a student might elect not to admit wrongdoing—this is a difficult decision that we help students make.
There is no single right way to respond to an allegation of AI misconduct. That said, there is a right way for you. We will help you find that proper response.
Call the LLF National Law Firm Student Defense Team at 888-535-3686 or contact us onlineto discuss your circumstances and how we may help you overcome this challenge.