Resources
A copyable AI use disclosure statement template for students: which tools to name, which tasks to describe, what stayed human, and which records support it.
Open core guideA disclosure statement turns hidden AI assistance into reviewable context. It protects students when a detector flags their work, gives instructors evidence instead of suspicion, and matches the direction of most academic-integrity policies: the question is rarely whether any tool was used, but whether substantive assistance was declared.
Name each AI tool you used and, where meaningful, its version or mode — for example a PDF reading assistant, a grammar checker, a translation tool, or a writing agent. Vague phrases like "I used AI" give reviewers nothing to evaluate; a named tool with dates of use does.
Describe what each tool did, at which stage: asking questions about assigned readings, organizing quotations, outlining, checking grammar, or translating. Task-level disclosure is what lets an instructor distinguish reading support and language help from generated analysis or drafting.
State explicitly what remains your own work: the argument, the analysis, the structure, the final wording, and the verification of every quotation and citation against its source. This is the half of a disclosure statement students most often omit, and it is the half that establishes attributable authorship.
AI Use Disclosure — [course, assignment, date]. Tools used: I used [tool name and version] between [dates]. Tasks: I used it to [describe tasks, e.g., ask page-cited questions about assigned readings; organize quotations by theme; check grammar in my final draft]. What stayed human: the [argument, analysis, structure, and final wording] are my own, and I verified every quotation and citation against the original source. AI was not used for [tasks the assignment prohibits]. Records: I can provide [reading notes, drafts, generation records, and an integrity review report] on request. Signed: [name, date].
ChatPDF keeps generation records with per-answer credit receipts while storing only a hash of each question, so a reading history exists without exposing your prompts. Integrity Review saves each run's input text, report, and credit receipt in a deletable review history. Writer Agent sessions keep a recoverable writing history. These records document assistance; they do not judge it.
This template is a starting point, not legal advice or a substitute for your institution's policy. The assignment instruction controls what must be disclosed and what is prohibited, so align the statement with the course policy, your institution's academic-integrity rules, and any funder or publisher requirements before submitting it.
Whenever the assignment or institution requires one, and whenever substantive AI assistance shaped the work. If the policy is silent, disclosing reading, organization, and language help is the safer and increasingly expected practice.
Policy decides how disclosed assistance is graded, but disclosure itself is protective: it converts a detector flag from an accusation into a documented, reviewable question, and most integrity processes treat undisclosed use far more severely.
Follow the assignment's categories. Many policies treat routine grammar and language support differently from drafting or analysis, but naming the tool and the task costs one sentence and removes the ambiguity.
It is written to be copied and filled in, but it is a starting point, not legal advice. The assignment instruction and the institution's policy control what must be disclosed, so adapt the bracketed fields to them.