University library hosting AI detection workshops for students

Our library launched a drop in service this semester where students can voluntarily run their own paper through our AI detection software before submitting, just to check for false positive risk, especially useful for our large ESL and international population but open to anyone. Its opt in, totally confidential, we dont report results to instructors or keep records tied to student names, its purely a self check. Uptake has been way higher than I expected, we’re booked most weeks students can drop in. The interesting wrinkle is we’ve had a few students come in genuinely anxious after getting flagged even though they insist, and I believe them, that they wrote everything themselves, and we dont really have great next steps to offer beyond here’s a screenshot for your own records if this ever comes up. Anyone else running something like this and if so what do you actually tell a student in that situation, flagged but genuinely innocent, beyond just documenting the result?