Wake County Student Challenges AI Detector Accuracy in School

A 15-year-old freshman at Green Hope High School in Cary, North Carolina, is now at the center of a growing debate over artificial intelligence detection in K-12 education. Eleanor Canina received a zero on an English I assignment about the first act of Romeo and Juliet after her teacher flagged the work as AI-generated.

The grade appeared on her digital gradebook alongside perfect scores from other classes. It came with a note from the teacher reading "evidence of AI, Please redo." Canina and her mother, Stacy De Coster, say the accusation was unfounded and the school's reliance on AI detectors was inappropriate.

According to GovTech, the teacher ran the assignment through three different AI assessment tools. The results showed likelihoods of 62%, 75%, and 87% for AI generation or significant AI assistance. Those percentages became the basis for the failing grade.

De Coster, a sociology professor at NC State, immediately pushed back. She suggested the teacher compare the writing with her daughter's past work and review the Google Doc history to verify the writing process. Instead, the teacher stood by the detector results.

The situation highlights a broader problem in education. AI detectors have demonstrated high rates of false positives across multiple studies. They often flag work from non-native English speakers, students with unique writing styles, or even work that simply uses certain sentence structures. The tools are notoriously unreliable (which should be obvious to anyone who has tried them).

Wake County Public School System did not directly address the family's accusations due to student privacy rules. In a statement, the district acknowledged that AI use in education is "a new and rapidly evolving area." They emphasized their obligation to ensure student work is evaluated fairly and consistently.

The district's position reflects a wider tension. Schools want to prevent cheating, but the tools available for detection are fundamentally flawed. This creates a catch-22 where teachers feel pressured to use unreliable technology because they lack better alternatives.

In 2024, the North Carolina Department of Public Instruction issued guidelines on AI use for public schools. The guidelines explicitly advise caution with AI detectors. They state that detectors should never be used as the only factor when determining if a student cheated. The Wake County teacher's approach appears to have violated this guidance.

Canina herself is no fan of AI for assignments. She recognizes the dangers of students using it to avoid doing their own work. "It's stopping people from thinking freely," she told The News & Observer. "Using it as a quick excuse to get out of doing work isn't going to help anyone in the long run."

Her position adds complexity to the case. This isn't a student defending AI use. It's a student arguing that the detection tools are too inaccurate to be trusted. The distinction matters because it removes the appearance of hypocrisy from her position.

The teacher's situation also adds context. De Coster noted that the English I teacher had resigned earlier in the semester. A long-term substitute is watching the class, but assignments are being graded by other teachers who don't know the writing styles of the students. This creates conditions where false accusations become more likely.

When a teacher doesn't know a student's baseline writing ability, they lose the most reliable detection method available: human judgment based on familiarity. AI detectors then become the default, despite their known limitations.

The physical reality of this situation is frustrating. Students now face the possibility of failing grades based on algorithmic probability scores. A 62% likelihood of AI use becomes a zero. The threshold for "proof" in education has shifted from evidence to statistical probability.

Canina and De Coster are waging this battle to protect other students from being falsely accused. They argue that Wake County teachers shouldn't rely on AI detectors that have been known to inaccurately flag work as AI-generated. The stakes extend beyond one student's grade.

The case raises questions about due process in academic settings. What evidence is required to accuse a student of cheating? What appeals process exists when AI detectors flag work? Schools have not universally answered these questions.

Technology companies selling AI detection tools have faced criticism for their accuracy claims. Some have admitted their tools are not reliable enough to make definitive determinations. Yet schools continue to deploy them as enforcement mechanisms.

This incident will likely not be the last. As AI becomes more prevalent in education, more students will face similar accusations. The question is whether schools will develop better verification processes or continue relying on flawed detection technology.

Whether Wake County revises its AI detection policies remains to be seen. The real issue is whether educators will accept that probability scores are not proof. Until that happens, students like Canina will keep fighting false accusations.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn