This study assesses the performance of 20 Generative AI (GenAI) models in answering multiple-choice questions within an introductory economics course at a business school. Contemporary GenAI models generally surpass the performance of average business school students, with paid subscription models demonstrating superior performance compared to free alternatives, suggesting that access to paid premium models could confer substantial advantages to students from more affluent backgrounds in assessments that allow external resources. Thus, traditional closed-book exams may present the fairest method for assessing student performance in introductory courses. The findings also highlight the potential of reliable GenAI models as cost-effective, round-the-clock tutoring tools for students in introductory courses, offering affordable alternatives to human tutoring for students.