Chris Rosser, First Year and Transfer Experience Librarian at Oklahoma State University’s (OSU) Edmon Low Library, used DocuMark in his Spring 2026 honors seminar, titled “They Wouldn’t Put It on the Internet If It Wasn’t True! Information, Society, Trust, and Technology in a ‘Post-Truth’ World”, and found it a powerful pedagogical tool that deepened student engagement with the writing process while meaningfully reducing student anxiety about being falsely accused of AI misuse.
In this honors course, Rosser asked students to critically examine concepts of surveillance, data ownership, and transparency.
DocuMark was found to be more than a process transparency tool supporting student work. In this course, it became a topic of active inquiry, as students analyzed the trade-offs of process transparency while experiencing them in real time.
Co-author of Generative AI and Libraries: Claiming Our Place in the Center of a Shared Future (American Library Association, 2025). Has developed significant expertise in gameful course design and AI-empowered instruction, teaching students not to fear or avoid AI but to engage with it transparently, critically, and purposefully.

Rosser integrated DocuMark into a signature assignment: “Watching and Being Watched: Academic Integrity, Surveillance, and AI.” Students opted into one of two writing paths.
Those in the DocuMark group used the platform to compose and revise a critical essay on process transparency, with DocuMark logging their writing sessions, revisions, and composition patterns throughout. A comparison group completed the same essay using their standard workflow, without DocuMark.
What made this pilot exceptional was its framing. DocuMark was not simply a tool students submitted work through — it was an object of inquiry in its own right. Students wrote about the trade-offs of process transparency while experiencing those trade-offs in real time.
They engaged with Shoshana Zuboff’s work on surveillance capitalism, scholarship on intellectual privacy, and Foucault’s theory of the panopticon, and then examined DocuMark through those same lenses, in their essays and in a structured class discussion.
My initial expectation was that DocuMark would function primarily as an integrity tool — a way to verify process. What I discovered was that it functioned better as a pedagogical tool — a way to make the surveillance conversation concrete for students. That was more valuable than I anticipated, and more interesting.
What Rosser found most valuable was not the summary scores but the granular process record beneath them, showing the session-by-session writing arc.
Seeing when a student wrote, how long each session lasted, where revisions clustered, and how early thinking related to the final argument gave him interpretive depth that a finished essay alone cannot provide. “Not the scores,” he observed, “The process.” That data made his feedback more targeted and his conversations with students more grounded: rather than responding to what a student had produced, he could respond to how they had worked: more interpretive, more contextual, and more genuinely useful than a simple assessment of a finished product.
The pilot also clarified what Rosser sees as DocuMark’s most promising long-term application: scaffolded writing instruction. Rather than deploying the tool at a single, high-stakes submission, he envisions requiring DocuMark reports at multiple draft stages: after an outline, after a first draft, after revision, so that conversations about process become iterative and formative rather than retrospective and evaluative.
DocuMark rendered process work legible, enabling me to provide formative feedback and design targeted interventions. Crucially, it shifted our classroom conversations about academic integrity from prohibition to empowerment, from ‘don’t cheat’ to ‘own your process.
“If I were teaching writing, I would have all the drafting done in DocuMark.”
One of the pilot’s most significant findings concerned student anxiety about false accusations of AI misuse, a documented challenge across higher education. A 2026 survey of 1,054 full-time UK undergraduate students by the Higher Education Policy Institute (HEPI) found that 53% fear being falsely accused of using AI, including students who have not. AI detection tools with high false-positive rates contribute significantly to this anxiety.
A post-pilot survey asked students two questions about their anxiety regarding false accusations of AI misuse: how they felt before using DocuMark, and how they felt during the pilot. Before using DocuMark, 100% of respondents reported at least some concern, either slightly or moderately worried. During the pilot, that figure fell by 37.5 percentage points, with 37.5% of students reporting no concern at all.
Oklahoma State University DocuMark Pilot User Survey | April-May 2026
One student articulated the shift directly in class discussion: without DocuMark, she felt exposed to AI detection tools that could flag her work incorrectly, with no process record to defend herself. With DocuMark, she knew exactly how her writing was being documented, and that knowledge was reassuring, not threatening.
The data reflects a key insight: DocuMark does not introduce surveillance into the writing process. It makes existing tracking legible and purposeful. Students knew their process was being documented, and that knowledge was protective rather than threatening. The record of their work became a resource they could point to, not a risk they had to manage.

Student feedback from the pilot is directly informing DocuMark’s ongoing product development.
The OSU pilot demonstrated that DocuMark’s value extends well beyond documentation. Students in the course offered detailed, analytically sophisticated responses about their experience, writing about consent, data ownership, the difference between transparent and opaque surveillance, and what it means to write authentically in a monitored environment.
These were not peripheral concerns; they were the heart of what the course set out to teach. DocuMark made that teaching tangible. As Rosser observed, the students’ reflections were precisely what the course was designed to produce, and the tool made them possible in a way that no reading or discussion alone could replicate.
The tool helped prepare students for a future in which transparent AI use is a professional expectation.”
DocuMark is Trinka AI’s writing process tool for higher education, designed to support students and instructors throughout the writing process, not just at the moment of submission.
Rather than evaluating a finished document, DocuMark makes the writing process itself visible: logging sessions, revisions, pacing, and composition patterns as students write, and generating a process record that instructors can review alongside the final work.
This gives instructors richer, more interpretive insight into how students think and write, enabling formative feedback that a finished essay alone cannot support.
For students, DocuMark provides a transparent record of their own process — documentation that builds confidence and replaces the anxiety of opaque AI detection with the assurance of a clear, reviewable account.
Trinka AI is compliant with FERPA, ISO 27001, SOC 2 Type 2, and the EU AI Act, and it does not use student data to train AI models.
To learn more or schedule a pilot conversation, book a demo