Brown University professor suspects AI cheating after midterm average drops from 96% to 48.6%

A Brown University professor suspected his class used AI to cheat after a take-home midterm exam averaged 96%, significantly higher than expected. In response, he administered an in-person final exam, which averaged 48.6%, a dramatic drop that he attributes to the absence of AI assistance. The professor criticized the university's response to the alleged cheating as "meek," highlighting ongoing tensions around academic integrity in the age of generative AI. The incident, reported by Emma Whitford of Inside Higher Ed, underscores the challenges educators face in detecting and addressing AI-assisted cheating. It also raises questions about assessment design and institutional policies on academic dishonesty.

Global Impact

The incident has limited global impact but is significant within the education sector. It highlights a systemic challenge: as AI tools become more accessible, traditional assessment methods may lose validity, forcing a rethinking of how learning is measured.