The California state bar has petitioned the California Supreme Court to lower the passing score needed for the February bar exam, a concession to technical difficulties that impacted exam takers. This notable petition, filed late Tuesday, highlights unresolved issues pertaining to the bar’s use of artificial intelligence (AI) in exam question development.
The bar confirmed publicly for the first time the involvement of OpenAI’s ChatGPT in creating some exam content. Specifically, AI was tasked with crafting certain questions on both a practice exam in November and the actual February bar exam. Despite the California Supreme Court’s inquiries, the bar did not explain why its test-writing vendor failed to deliver the full set of 200 questions or the rationale behind employing AI to fill this gap.
Legal professionals following this development may be interested in how the use of AI in high-stakes testing could influence future legal examinations and the standards of entry into the profession. The broader implications of relying on AI tools in such contexts, particularly for fairness and evaluation metrics, remain under the spotlight.
For more complete coverage, refer to the detailed article available through The Recorder.