Iowa students ask for a say on AI in the classroom: the state rolls it out while they already use it and fear it

By Daily Nonpareil (Des Moines Register / USA Today Network) · October 20, 2026.

Iowa high school students are demanding a say in decisions about how AI is used in their schools. The Des Moines Register reports the story in a piece published by the Daily Nonpareil. It centers on Dylan Dornack (Creston High School, president of Iowa FBLA), Maxwell Landers (Winterset High School) and Avery Corso, a Valley High School graduate now studying at the University of Chicago.

This summer, Dornack and Landers joined 96 other students from all 50 states at a Student Senate on AI Policy held in Boston. Over three days they debated how these tools should be used in classrooms before voting on a K-12 framework called the "STUDENT FIRST Act." According to the account, the proposal allows certain uses, such as brainstorming, studying and editing, with teacher permission and disclosure. It also calls for AI literacy classes. And it states that AI cannot decide a grade or a punishment on its own: a student accused of cheating with it would be entitled to human review.

The article provides the following context:

- More than 96,000 Iowa elementary students used Amira, a state-sponsored AI reading tutor, during the 2025-26 school year. The program gives children feedback as they read. - Iowa first spent $3 million in federal pandemic relief funds on Amira. Governor Kim Reynolds then approved another $2.5 million. The state wants to offer it free to up to 110,000 elementary students through July 2028. - According to state officials, those who used it at least 20 minutes a week made greater reading gains. Parents, however, have raised concerns about voice data and about how the tutor compares with help from a person. - An August Common Sense Media survey indicates that 70% of teenagers use AI for homework. A 2026 study of more than 4,000 high school students concluded that generative AI had not increased the prevalence of cheating. - Starting in July 2027, AI chatbots used by minors must disclose that they are not human and follow restrictions on sexual content, emotional dependence and self-harm conversations. The article does not specify the exact legal status of this rule. - On the jobs front, a June survey found that 55% of teenagers fear AI will hurt their chances of finding work. Another survey found that 65% of current students believe AI takes opportunities away from entry-level workers.

The piece also includes testimonies that illustrate the tension. Landers says he uses AI to turn notes into flashcards and to critique his essays. Corso recounts that in a calculus class many classmates did their homework with AI and then failed the exam without it. And Dornack says his school blocked ChatGPT on district computers, to little effect, and that he sometimes finds it "hypocritical" that teachers use AI themselves.

Our take: this story isn't about cheating, it's about governance. There is a clear mismatch between a state investing millions to roll out AI in elementary schools and districts that, according to the article, are still debating whether to let students use it. Meanwhile, students are already using it. Banning without explaining, like the block on district devices, stops nothing: it shifts use to phones at home and removes the chance to teach when AI helps and when it replaces learning.

Corso's story is the best clue. If AI does the homework and the student fails the exam, the problem isn't that the tool exists. It's that the assignment doesn't measure what the student knows. The schools that win will be those that redesign assessment (more oral, more in-person, more process-based) instead of policing students. That fits a thesis we champion at Zendoric: the teacher who orchestrates AI gains value, and the one who only transmits content loses it.

It's also worth being cautious with the data. Amira's reading gains come from state officials, who are an interested party. The finding of no increase in cheating comes from a single study, and the 70% usage figure comes from a survey. None of the three yet allows us to say that AI improves learning in general. The criticism from Jodi Carreon of Schools Beyond Screens, who calls for a moratorium like the one adopted in New York schools, starts from the same point: the technology is new and deployment is moving faster than the evidence.

The data point that weighs most in the medium term is the one on jobs. More than half of teenagers fear for their working future. In general, it is noted that AI's impact shows up first in entry-level positions that never get created. In the short term, that anxiety is reasonable and schools cannot ignore it. In the long term, the AI literacy these students are asking for is the best preparation for an economy where what will change is which work matters, not whether there is work.

The most interesting thing about the Student Senate is that it points to a replicable model: giving students a seat at the table, with clear rules on privacy, human review and transparency. It is a cheap and effective way to get rules followed, because they were agreed on by those who live with them. The adults designing these policies would do well to take it seriously, and not just out of courtesy.