AI/ChatGPT Will Not Be Used To Generate OLA or Exam Content
By Brent Wagner, MD, MBA, ABR Executive Director, and Matthew B. Podgorsak, PhD, ABR Board of Trustees Chair
2024;17(2):4
The accelerated evolution of generative artificial intelligence (AI), including ChatGPT, has revealed not only a new horizon of seemingly boundless applications but also myriad concerns, as large language models produce content that is often indistinguishable from human speech or thought.1 Citing the use of source material as inputs to such models, institutions, including the New York Times, have recently sought to protect their copyrighted work through the judiciary.2
The ABR has three specific challenges related to AI. First, we want to ensure that questions administered as part of our Online Longitudinal Assessment (OLA) are not introduced into the public domain via generative AI tools. Second, acknowledging that there may be significant limits to copyright protections of content that is not produced by humans, the ABR has an interest in reminding volunteers that writing questions for ABR exams is a human endeavor. Last, if an ABR question were written by generative AI software, it may violate the copyright protections of the original authors of the work.
Individuals engaged in OLA agree, as a condition of their participation, not to distribute or share the questions and related content outside of the interface. ABR volunteers are required to refrain from using generative AI when developing exam content. In this video, Board of Trustees Chair Matthew Podgorsak, PhD, shares more details about these requirements.
1 Mahowald, K, et al. “Dissociating language and thought in large language models: a cognitive perspective.” arXiv preprint arXiv:2301.06627 (2023). 2301.06627.pdf (arxiv.org)
2 Grynbaum M and Mac R. “The Times Sues OpenAI and Microsoft Over A.I. Use of Copyrighted Work.” https://www.nytimes.com/2023/12/27/business/media/new-york-times-open-ai-microsoft-lawsuit.html
