Turnitin adds an AI writing score to its plagiarism reports, flagging submissions it believes were AI-generated. Humanums takes the opposite approach: students capture behavioral signals while writing and submit a certificate as proof of original work. One accuses. The other empowers.
Turnitin scans submitted text and flags potential AI content with a percentage score. Humanums captures the writing process and lets students prove their work is original through behavioral certification. The two tools represent fundamentally different models of academic integrity.
Turnitin runs after submission and produces a score the student cannot control or contest with evidence. Humanums runs during writing and gives the student portable proof.
Turnitin has acknowledged that its AI detection scores should not be the sole basis for academic integrity decisions. Humanums provides behavioral evidence that can be independently verified.
Turnitin is built for institutions to police students. Humanums is built for students to prove themselves.
Turnitin added AI writing detection to its existing plagiarism platform. When an instructor submits student work, Turnitin returns a percentage score estimating how much of the text was AI-generated. The score appears alongside the traditional similarity report.
Turnitin itself has published guidance stating that AI detection scores should not be used as the sole evidence for academic integrity decisions. The tool is a signal, not a verdict.
The adversarial model is the core issue. A student submits work, the institution runs a scan, and a score appears that the student has no way to independently verify or dispute with their own evidence. If the score is wrong, the student is left arguing against a black box.
Documented bias against non-native English speakers makes this worse. Students writing in a second language produce prose patterns that classifiers are more likely to flag, even when the work is entirely original. The tool penalizes the students who can least afford a false accusation.
Humanums flips the model. Instead of scanning text after submission, it captures behavioral signals while the student writes: keystroke timing, pause patterns, revision behavior, paste ratio, and session distribution. The student then certifies the work and submits a verification link alongside the paper.
The instructor can click the link and inspect the behavioral evidence. The proof comes from the student, not from an institutional scan. That changes the relationship from adversarial to collaborative.
Yes. Turnitin handles plagiarism detection, which is a separate problem from AI authorship. An institution can continue using Turnitin for similarity checking while adopting Humanums as the mechanism for students to voluntarily prove human authorship.
The two tools address different questions. Turnitin asks: is this copied? Humanums asks: was this written by a human? They are complementary, not competing.
Turnitin itself has cautioned that AI scores should not be used as sole evidence for integrity decisions. Independent research has shown higher false positive rates for non-native English speakers.
Yes. Students write in the Humanums editor, certify their work, and submit both the paper and the verification link. Turnitin can still run its plagiarism check on the submitted text.
No. Students use Humanums independently. The verification page is public and requires no software, account, or LMS integration for the reviewer.
Humanums captures the writing process and issues a certificate students can submit alongside their paper.