Grepp's "Monito" Adopted by 21 Companies to Prevent Smart Glasses Cheating

Four areas of glasses photographed before the test; proctors verify them directly

Grepp said on October 28 that 21 companies had adopted the “Precise Glasses Verification” feature of its AI proctoring solution, Monito, just two months after its launch.

Grepp added the feature in August to address cheating involving smart glasses. It took into account that smart glasses look similar to regular glasses and are difficult to distinguish with the naked eye, and that AI could make an incorrect judgment if used alone.

Before the test begins, test takers wearing glasses use the Monito application to photograph four areas of their glasses: the front, side, inner side, and both temple tips. The AI can take the photos automatically, or test takers can take them themselves.

Grepp said on the 28th that 21 companies had adopted the "Precision Glasses Verification" feature of its AI exam proctoring solution, "Monito," within two months of its launch. Grepp.

Proctors can view each test taker’s glasses photos in the online proctoring room and record any notable details. If a situation raises suspicion during the test, proctors can compare the glasses with the photos taken beforehand and determine for themselves whether they are smart glasses. The data can also be reviewed again after the test. To date, 21 companies have adopted the Precise Glasses Verification feature, including companies in IT and semiconductors, finance, manufacturing, automotive, retail and commerce, as well as state-owned enterprises, public institutions, and broadcasters.

Grepp is also developing a feature based on a VLM (vision-language model) that analyzes behavioral context and alerts proctors to signs of suspected cheating, including cases where test takers switch to smart glasses during a test. The feature is targeted for an update in November.

Grepp CEO Lim Seongsu said, “From the outset, when designing Monito, we took great care to ensure that test takers would not be unfairly judged because of errors in the technology. Going forward, we will add features that interpret behavioral context to support accurate decisions by proctors and build a testing environment that both companies and test takers can trust.”

This content was produced with the assistance of AI translation services.