Why it matters when you are writing
It is broader than plagiarism, which is only the attribution part. The same principle covers fabricating or selectively reporting data, misrepresenting your methods, listing an author who did not contribute or omitting one who did, reusing your own work without disclosure, and undisclosed conflicts of interest. What unites them is that a reader’s ability to assess the work depends on the account being accurate.
Policies differ in ways that actually matter, and the differences are local. Two universities in the same city can take opposite positions on whether a proofreader may change your sentences, on how much of a taught assignment may be discussed with peers, and on what use of AI assistance must be declared. Reading your own institution’s policy once, properly, is a better investment than any general advice — including this page.
Where the rules are genuinely unclear — and around AI tools they often still are — disclosure is the move that holds up. A short note stating what assistance you used and for what turns an ambiguity into a documented decision. So does keeping a record of your process: drafts, notes, dated versions. If a question is ever raised, the writer with a trail is in a different position from the writer with only an assertion.
