A statistics professor works tirelessly to catch students cheating on his exams. He has particular routes for his teaching assistants to patrol, an elevated chair to ensure an unobstructed view of all students, and even a video recording of the exam in case additional evidence needs to be collected. He estimates that he catches 95% of students who cheat in his class, but 1% of the time he accuses a student of cheating and he is actually incorrect. Consider the null hypothesis, "the student is not cheating." What is the probability of a Type I error?