Wearable AI puts AI-proof exams in serious jeopardy, researchers say

AI glasses, earbuds, watches, and other devices are making AI-related cheating in exams more difficult for universities to detect and prevent.

Researchers describe wearable AI as a second wave of disruption for assessments since the launch of ChatGPT in 2022. Deakin University researchers have called it a “wicked problem” that exacerbates the issues that educators are facing with students using AI to cheat.

These wearable AI tools include glasses, earbuds, rings, pendants, watches, and hearing aids. Since they resemble ordinary accessories that a student might wear, they enable students to cheat undetected.

Institutions may say their exams are AI-free without possessing a reliable basis for claiming so

Researchers say that cheating can take the form of information retrieval, scene interpretation, and conversation coaching. The devices are controlled silently through the student’s movements, and exam invigilators can be completely unaware of them.

In the journal Assessment and Evaluation in Higher Education, the Deakin University researchers have described these tools as the final nail in the coffin for AI-free exams, despite attempts to stop its use through supervised performances, interactive oral assessments, and invigilated exams.

Some universities have banned wearable AI, but researchers argue that this is futile and could discredit the institution; they may say their exams are AI-free “without possessing a reliable basis for claiming so”.

Thomas Corbin, the lead author of the research, said that assessors were already having to look into students’ usage of words like “therefore” or em dash punctuation in their work, as these are common pieces of evidence for generative AI usage.

Universities have a big decision to make: do they continue trying to ban AI use, or do they make exams a thing of the past?

Assessors now have to do even more detection work and may be placed in very difficult situations, particularly when students wear hearing aids or religious head coverings. According to Corbin, this is “a world that no teacher wants to live in”.

The researcher suggested that the solution is not to try to ban AI usage in vain but instead use a different kind of assessment that won’t be affected by it. This may lead to exams being stopped altogether.

Wearable AI, according to Corbin, can also be “educationally virtuous” and “can overcome a lot of barriers to entry … into higher education [through functions] like translating for international students or overcoming hearing barriers”, which is another reason why banning them may not be the answer.

Universities have a big decision to make: do they continue trying to ban AI use, or do they make exams a thing of the past?