A Teacher Was Fired for Refusing to Use AI Grading Software. Her First Amendment Lawsuit Is Now a National Story

Although someone has covered it with a piece of painter’s tape, Hannah Carter’s name is still on the door of the classroom where she spent eleven years. Back in November, a student sent her a picture of it. It is stored on her phone. Talking to her gives the impression that she didn’t think any of this would go as far as it has, and perhaps that’s why it did.

Carter’s refusal to use an AI grading platform required by the district for her juniors’ argumentative essays led to her termination last spring. For a semester, she had reluctantly used it. She stopped then. She claimed that the program was “flattening” her students’ voices, rewarding formulaic writing, and, in her opinion, penalizing the children who took chances in an email to her principal. She was instructed to use it nonetheless by the principal. She didn’t. She lost her job a few weeks later.

The legal argument her attorneys are making is what makes her case unique and the reason it has attracted attention from media outlets that don’t typically cover small-town school board disputes. They are portraying her refusal as an act of professional conscience, similar to a teacher declining to assign a book that is prohibited, and as protected speech. It’s a new theory. In the past, courts have granted school districts broad discretion over curriculum and evaluation. However, judges appear to be at least open to listening because of the current situation, the rapid advancement of AI in classrooms, and the growing discomfort among educators.

According to Carter’s district, she was just disobedient. Teachers abide by the rules. Policy was the software. The story is over. Legally speaking, that is the better case, and it might win. However, it’s difficult to ignore how meticulously she recorded everything when reading her deposition, which was made public in February. For example, the student sobbed when the algorithm gave her a 2 out of 6 on a piece about her grandmother’s dementia, and the third-period boys began writing in the precise cadence the software seemed to prefer. She maintained a folder. She had receipts with her.

Here, the larger context is important. During the 2024–2025 school year, about 85% of teachers used AI tools. Districts, under time and financial constraints, have been implementing automated grading platforms at a rate that surpasses any meaningful discussion about how they affect teaching. A portion of the zeal is merited. Teachers do receive their hours back. Students frequently receive feedback more quickly. However, only 6% of educators think AI is more beneficial than detrimental to education; this percentage keeps coming up in surveys, and no one is sure how to handle it.

You begin to see that Carter’s case isn’t really about her as you watch this develop. It concerns a question that the nation has been avoiding: who gets to say no and what gets automated in a classroom? Efficiency will be emphasized by administrators. Parents will bring up equity. The one thing that still feels like craft is being given to the teachers, who are the ones who actually read the essays at eleven o’clock at night while sipping cold tea.

Fall is when her trial is scheduled to take place. People who are following the case feel that the conversation has already changed, regardless of the outcome. There is no end in sight for the software. However, it turns out that neither is Hannah Carter.