Detecting AI Plagiarism

Researchers at a large midwestern university have developed algorithms to detect when a submitted assignment was written by artificial intelligence $($AI$)$ instead of by the student. The researchers have shown that these algorithms correctly identified a non$-$AI $($student$-$written$)$ submission $98\mathrm{.}3\%$ of the time and correctly identified use of an $Al94\mathrm{.}7\%$ of the time. Note that students are assumed to not have plagiarized using an Al until evidence of Al use has been detected.

Describe what a Type II error would mean in terms of the "Detecting the AI Plagiarism" scenario