Whether or not a student is permitted to use artificial intelligence (AI) to create a written draft, their use of AI tech could be flagged by an AI detector—yes, even if they revise the AI’s output into their own words.
Hard as a student may try to alter an AI’s words, that AI may leave behind remnants that AI detectors pick up. A single sentence, repetitive use of a particular type of punctuation, or other possible sign of AI-generated text could expose the fact that a student used AI to write the text.
This issue is complicated. For one, students may be unclear whether they can use AI to write a draft, but then revise AI’s writing. Secondly, the platforms that might “flag” a student’s writing are known to have significant false-positive rates.
As nuanced as this issue is, universities nationwide sanction students in severe ways (and on a regular basis) merely because that student’s work was “flagged.” If you or your student finds themselves in such a situation, do not wait to contact the LLF National Law Firm Student Defense Team at 888-535-3686 or contact us online.
Are University Students Allowed to Use AI to Write a Draft, Then Revise?
The vast majority of universities, in America and elsewhere, have no single rulebook for AI use. Many of them have guidelines, but not a single, all-encompassing, non-negotiable book of rules.
Therefore, when someone asks, “Can I use AI to write a draft?”, the most reasonable answer is: Ask your instructor.
We can find individual instructors’ AI usage policies that:
- Allow students to use AI when drafting work: the University of Minnesota AI policy, for example, states that students may “Utilize Generative AI for initial draft work.” More specifically, the policy endorses AI as a tool to “assist with idea creation, brainstorming, and research gathering that offer new ways to write about a topic.”
- Explicitly prohibit students from using AI in the drafting process: Conversely, many policies are more restrictive of students’ AI use, even for drafting purposes. One University of Michigan policy, for instance, states that “Writing a draft of a writing assignment” is “not permitted in this course.”
At most schools, each instructor has their own AI usage policy. The details of such policies will determine whether a student can use AI in drafts and revise the work into their own words (without facing possible disciplinary action).
Can My School Have Evidence That I Used AI, Even If I Revised the Vast Majority of the Writing?
The initial question is whether your work could get flagged if you used AI to create a draft. When we say flagged, it generally means that an AI-detecting software found possible signs of AI-generated content in your work.
Some of the purported evidence that an instructor or other university official might cite when accusing a student of AI misconduct are:
- A result from an AI-detecting platform suggesting that some portion of the work was likely generated by artificial intelligence
- Any statement the student makes suggesting, or outright admitting, that they may have used AI in a prohibited manner
- Stark differences in the student’s work, whether that is a marked improvement in quality, a notable change in language or style, or other disparities that might indicate the involvement of an AI resource
- Witness statements suggesting that a student used AI in an unauthorized way
- Metadata contained within a submission, including but not limited to revision history (such as tracked changes), watermarks embedded by AI platforms, and the total editing time
These and other types of evidence are not always reliable. For instance, AI detectors are known to raise false positives. A student’s words might be taken out of context. Metadata may lack context, especially if a student copied and pasted their work from one platform to another before submitting the final document. A student might also significantly improve their writing through hard work, rather than using AI to assist them.
When our Student Defense Team represents students accused of AI misconduct, we often exploit the flaws in the supposed evidence of their guilt. This is just one of the many ways we step up for students who face serious allegations of academic misconduct.
Can My School Use the Result of an AI-Detecting Platform in a Disciplinary Case Against Me?
Your university’s policies will likely dictate whether results from an AI-detecting platform can be used as evidence in a disciplinary case. These policies can vary significantly, with a few examples including:
- Texas State University, which says that scores from AI detectors “cannot be the sole evidence of academic misconduct when submitting an Honor Code violation,” which suggests that the score could be used as evidence, so long as it is not the sole evidence suggesting academic misconduct
- The University of San Diego, which states that “AI detectors are problematic and not recommended” as the primary method of determining whether a student has engaged in AI-related misconduct
These policies are reflective of most. However, some universities allow instructors and other officials to rely more instrumentally on AI detector results to adjudicate suspected misconduct.
What Should I Do If I Am Facing Possible Discipline for AI Misconduct?
Any student accused of any form of academic misconduct should contact our Student Defense Team right away. Whether you are accused of using AI in an unauthorized way to create a draft or have been targeted by another type of accusation, you likely have much at stake.
Our Student Defense Team has resolved many high-stakes disciplinary cases. We even negotiate with universities’ general counsel when it is the appropriate strategy for the case at hand.
Some of the ways we might help you or your student are:
- Citing the well-known flaws in AI detector technology: Studies and anecdotal evidence both indicate high error rates by many platforms that purport to spot AI-generated writing. We will highlight this fact, as relevant, when leading your case.
- Encouraging a more comprehensive evaluation process: When evaluating a student’s case, the evaluators should consider the entire picture of evidence and facts. This picture may include the student’s prior displays of character, which may not jibe with the suggestion that they engaged in unethical behavior.
- Ensure you are ready for everything to come: Disciplinary proceedings can be daunting and fast-moving. Our team will bring some calm to a disorienting process.
Now is the time to get a defense you can trust. Contact the LLF National Law Firm Student Defense Team at 888-535-3686 or contact us online today.