Academic authorship, verified.

Proprietary method, in development. Details are shared under a confidentiality agreement.

Detectors guess. Teachers can't prove anything.

61.3%

A Stanford study found AI-writing detectors misclassified non-native English speakers' essays as AI-generated 61.3% of the time on average.¹

26%

A leading AI developer shut down its own classifier in July 2023 after it flagged AI-written text correctly only 26% of the time.²

Our approach

Detection tools estimate. We are building verifiable evidence of authorship, with a proprietary method.