Abstract: Researchers who study how artificial intelligence (AI) methods explain their decisions often discuss the field’s controversies and limitations. Some even contend that many publications offer little or no substantive contribution. In this talk, I illustrate the claim that explainable AI (XAI) is in trouble by describing problems with definitions, motivations, and evaluation practices. I link these problems to the field’s interdisciplinary nature and to inadequate scientific rigor. This analysis yields a set of open research questions and recommendations for avoiding these problems.
Bio: Rosina Weber is a Professor of Information Science and Computer Science at Drexel University where she advises students with interdisciplinary interests. With degrees in both Economics (B.A.) and Engineering (M.S., Ph.D.), she is a leader in explainable artificial intelligence (XAI) and case-based reasoning where multiple disciplines converge. Weber has spent more than two decades combining symbolic and neural methods to build use-inspired AI systems across biomedical, legal, military, and science-and-technology domains. Her research has been funded by NIH, DARPA, DHS, and international agencies, including projects such as DARPA POCUS-AI (improving model accuracy through XAI), the NIH NCATS Biomedical Data Translator (designing and explaining a reasoning agent), Sweden’s Vinnova-funded initiative to embed explanatory capabilities in deployed AI applications, and the DARPA ITM (aligning an interpretable model to decision maker’s attributes). Professor Weber has co-chaired multiple XAI workshops, delivered XAI tutorials, and taught AI to both computer-science majors and students from non-technical disciplines. Her scholarship appears in venues such as AI Magazine, Applied AI Letters, Expert Systems with Applications, AAAI, and ICCBR; her papers have earned best-paper honors, and she has received a Research Excellence Award. Beyond academia, she has been featured on Good Day Philadelphia and NBC’s Nightly News with Lester Holt, as well as podcasts. She is an elected member of the AAAI Executive Council and also a member of AAAS, AWIS, and ACL.