Why St. Paul Public Schools needs clearer AI detection rules — now

Posted
Central High School, St. Paul

This story was produced as part of ThreeSixty Journalism’s 2026 Opinion and Commentary Workshop for youth, in partnership with MinnPost, Minnesota Reformer, Minnesota Star Tribune and Sahan Journal. ThreeSixty Journalism equips young Minnesotans to be sense makers, truth tellers and bridge builders for an equitable society through multimedia journalism. 

The night before my last day of sophomore year, I checked the score for an English essay that would significantly affect my final grade. I was expecting the essay to score high, given I had earned a 98% in the class up until then. Instead, I saw a D as my grade and a 0 on my essay. The note below it read, “Flagged as 32% AI. You need to rewrite.” 

I went to see my teacher first thing in the morning, offering to rewrite sections with pen and paper on the spot, despite not having used AI for any of them. My teacher said she knew I would, but that it wasn’t necessary. She had seen my writing evolve over the years and knew my style. She believed my work was original, expressing prior concerns about Turnitin, an AI detection tool used by the St. Paul Public School (SPPS). 

Until this incident, I’d earned As for my writing. My teacher warned I might need to change my style to avoid getting flagged for AI in the future. This prospect was upsetting. This teacher understood. But, what if she hadn’t? The problem lies with the limitations detection tools and a lack of clear, consistent districtwide guidelines on how to handle such situations. Administrators, not students, should fix the problem, but there appears to be an unfortunate lack of urgency. 

In September 2025, Phil Wacker, SPPS high school curriculum supervisor, acknowledged Turnitin’s limitations, particularly for students whose first language is not English, according to St. Paul Publishing. At the time, nearly 30% of the district’s students were English learners, making this especially problematic. There still isn’t an official school board AI policy, but Wacker said “those conversations are happening.”  

The SPPS Board of Education recently announced its goal of adopting an AI policy by the end of the 2027-28 school year. Details remain vague, with the board stating that the policy will include “expectations” for how SPPS prepares students and staff “to use, question, evaluate, and create with AI.”  

But the district needs to move faster because the time spent debating a policy means that, for my class, the “conversation” will take our entire high school career. Meanwhile, some students use AI and have found ways to avoid detection. AI use should be regulated, but tools such as Turnitin are limited in restricting its use. 

Turnitin assigns a score based on the percentage of suspected AI sentences, measuring “burstiness” — spontaneity of word choice — and sentence length. The company claims human writing “tends to be inconsistent and idiosyncratic.” Ironically, my essay was about how high-school writing structure strips individuality from students’ work. As my flagged paper noted, students are taught to follow a specific sentence structure that can read as robotic and unoriginal. And final essays are heavily edited. If you account for these two aspects of high school writing, you can anticipate higher false positive rates than Turnitin’s claim of 1%. Students are expected to follow rigid writing structures, but when executed correctly, they risk AI accusations — a double standard that terrorizes students.  

My teacher understands Turnitin’s flaws. While it flagged my writing, my teacher said she suspects some AI use in the class went undetected. Leaving punishment to a teacher’s discretion risks bias and arbitrary protocols. Without a concrete policy to guide grading, SPPS is allowing inequity. 

The Minnesota Department of Education’s guidelines for AI use state that decisionmakers must understand how AI models work to anticipate limitations, problems, and risks. The guidelines acknowledge risks such as bias and problems with academic integrity; but they do not provide specific steps for teachers to investigate a student’s disputed AI-detection result. 

According to the SPPS “Uses to Avoid” section of its AI Best Practices for Educators document, AI assessment should not be used for absent human verification. However, it does not list the verification steps for teachers to reference. Turitin’s guide for using its detection tool states: “It takes human judgment in conjunction with an organization’s application of its specific academic policies to determine whether academic misconduct has occurred.” For Turnitin to be an effective tool, SPPS needs to come up with “specific academic policies” that help teachers make equitable judgments.  

In this instance, my teacher concluded that it had been a false positive. I earned a 59/60 on my essay after she ruled out AI use. While my work, reputation and GPA remained intact this time, I still worry about students unfairly accused of using AI. 

In August 2026, KSTP reported that the Royalton and St. Cloud districts developed AI policies, while other districts are “waiting to see how their peers handle AI.”  

SPPS needs to stop waiting for examples. AI is developing so fast that the board’s plan may be irrelevant by my graduation in 2028. Students, and our teachers, deserve a policy to follow that removes ambiguity and addresses the moment. SPPS doesn’t have the luxury to observe how their peers handle AI. The district needs to take initiative and provide clarity now. 

The post Why St. Paul Public Schools needs clearer AI detection rules — now appeared first on MinnPost.

Community Voices, Opinion