In 2025, the U.S. Department of Education’s Institute of Education Sciences selected TrueMark for its Small Business Innovation Research program, which funds early research on promising education technology. The award paid for TrueMark to build a personalized AI Copilot for student writers, and for an outside research firm, Quantily, to study how it worked in real classrooms.
From February to May 2026, eight teachers and their students used TrueMark at four schools: public, private, and charter, in Nevada, New Jersey, and New York. Quantily’s researchers interviewed the teachers, surveyed students at the start and end of the pilot, and analyzed writing data from about 400 students. The quotes below come from Quantily’s recorded interviews, although Quantily redacted the interviewees’ names before handing TrueMark their findings.
measured writing skills improved
follow-on research grant from the U.S. Department of Education
Every teacher in the study noted that academic integrity was their primary reason for using TrueMark. They were able to define what original work looked like on an assignment-by-assignment basis, and provide their students with guardrails that encouraged original work.
teachers named academic integrity as their primary reason for using TrueMark
Teacher, pilot study
Many of these teachers started out wary of AI in the classroom, but they also knew it was not going away. One teacher described where their school was:
Teacher, pilot study
In TrueMark, students write in a workspace that keeps a record of how every draft is written, providing proof of their learning process. When students do try to hand over the work to AI, in one teacher’s words, “TrueMark catches [students] every time.”
TrueMark’s ability to draw out students’ original work is what teachers prized the most, and what they wanted to stay the same as the tool continues to mature.
Teachers can also enable optional AI assistance for students on their assignments. Because TrueMark’s AI Copilot is based on the Socratic method, if enabled, it asks questions that push students to work out the answers themselves and develop their own voices. When students pushed it to do the work for them, it refused. One teacher called it “sassy.”
Students’ most common complaint was that the Copilot would not give them the answers they were looking for. The third-party research firm, as well as the teachers they interviewed, saw this as evidence that TrueMark was working as designed. Some students appreciated this:
Student, pilot study
Teachers liked that they no longer had to choose between pen/paper and the “Wild West.” They liked that TrueMark allowed them to limit the number of times students could chat with it, and that they could install custom guardrails depending on the assignment. Put together, these controls helped teachers define what appropriate AI use looked like.
Teacher, pilot study
Seven of the eight teachers in the study reported concrete signs that their students’ writing was improving over the course of the semester. The eighth never turned on TrueMark’s AI Copilot for their students. In other words, in the study’s limited sample, every teacher who used the Copilot saw their students’ writing progress, with these improvements carrying over to pen and paper writing assignments as well.
7 teachers used the AI Copilot and reported concrete signs that their students’ writing was improving
1 teacher never turned the AI Copilot on
The signs were specific. One teacher said the writing was the best they had seen in three years of teaching the subject. Another stopped handing out essay checklists.
Teacher, pilot study
Teachers also saw the gains hold when TrueMark was not in front of students. One noticed better organization on handwritten assignments, which students completed without the platform.
The writing data pointed the same way. The study tracked five writing skills across a student’s assignments: building a claim, supporting it with evidence, analysis and reasoning, organization, and language and style. Over the course of the pilot, students improved on all five. The largest gains were in analysis and reasoning and in the use of evidence.
Two patterns stood out. Students who completed more assignments in TrueMark improved more. And students who started with the weakest skills improved the most.
Largest gains
Largest gains
Students who completed more assignments in TrueMark improved more.
Students who started with the weakest skills improved the most.
Students noticed the difference themselves. In surveys at the end of the pilot, they agreed that TrueMark had improved their writing. About six in ten named the rubric checker – which allows students to get targeted feedback on their draft against their teacher’s rubric – as their favorite feature.
6 of 8
teachers said TrueMark helped them change how they teach
8 of 8
teachers found TrueMark’s core features easy to use
Six of the eight teachers said that TrueMark helped them change how they teach.
One was able to assign two to three more writing pieces each marking period. Two began grading to a higher standard on assignments where students had the Copilot and the rubric checker. One used TrueMark to simulate AP exam conditions.
All eight teachers found TrueMark’s core features easy to use. A number of them noted that it made learning more accessible: students that were stuck no longer had to wait for the teacher, and quieter students were able to ask the Copilot questions they would not ask out loud.
Teacher, pilot study
This small pilot study was not designed to be comprehensive. Its purpose was to learn whether TrueMark works in real classrooms and whether the early results justified a larger study and further product development.
Because it was not designed to prove cause and effect, there was no comparison group of students writing without TrueMark.
The study showed that students improved and that their teachers saw it. It did not show how much of that improvement TrueMark caused.
That is something that our next phase of research and development will seek to establish. In fall 2026, the Department of Education awarded TrueMark $1 million to continue this research. A larger study, using more formal research methods, will run during the 2027-28 school year.
