01
Continuous Active Learning
Developing and validating CAL for high-recall information retrieval and technology-assisted review.
Artificial intelligence · law · courts · healthcare
Research Professor, lawyer, clinical psychologist, and court-appointed Special Master focused on the trustworthy evaluation and use of artificial intelligence.
Maura R. Grossman is a lawyer, clinical psychologist, computer science research professor, and court-appointed Special Master whose career has spanned psychology, legal practice, artificial intelligence, and evidence-based decision making. She earned a B.A. in Psychology from Brown University and a joint clinical and research Ph.D. in Clinical/School Psychology from Adelphi University, practiced as a clinical psychologist and hospital administrator, and later earned her J.D., magna cum laude, from Georgetown University Law Center, where she was elected to the Order of the Coif.
Following seventeen years at Wachtell, Lipton, Rosen & Katz, Grossman founded Maura Grossman Law and joined the University of Waterloo, where she is a Research Professor in the David R. Cheriton School of Computer Science with a cross-appointment to the School of Public Health Sciences. She is also a Faculty Member of the Vector Institute for Artificial Intelligence. She has taught at leading law schools, including as a Visiting Professor at Duke University School of Law and an Adjunct Professor at Osgoode Hall Law School, where she developed and taught an interdisciplinary course bringing together Osgoode law students and University of Waterloo computer science students for joint seminars and mixed-team projects. She frequently serves as a court-appointed Special Master in complex federal litigation involving advanced technologies.
Grossman’s research focuses on the trustworthy evaluation and use of artificial intelligence in law and healthcare. She introduced the term Continuous Active Learning (CAL), demonstrated the effectiveness of the methodology it describes, and has applied it to technology-assisted review, systematic evidence synthesis, and the empirical evaluation of legal AI. Her current work addresses generative AI, AI-generated evidence, and rigorous methods for validating AI systems.
01
Developing and validating CAL for high-recall information retrieval and technology-assisted review.
02
Helping courts address deepfakes, AI-generated evidence, authentication, and responsible judicial use of AI.
03
Designing empirical methods for comparing review processes and evaluating AI under real-world conditions.
04
Applying CAL to systematic and rapid reviews in medicine and public health, including large-scale evidence synthesis.
Demonstrated that technology-assisted review could be more effective and more efficient than exhaustive manual review.
Read about the JOLT paperEstablished an autonomous, minimally tuned CAL protocol and a durable benchmark for high-recall review across diverse datasets.
Read about the paperDefined a common technical vocabulary for courts, lawyers, vendors, and researchers working with technology-assisted review.
Read about the glossaryExamined how generative AI challenges evidence, authenticity, litigation practice, and judicial decision making.
Read about the Duke Law paperProposed independent comparative evaluation in place of conventional validation built around uncertain recall estimates and arbitrary targets.
Read about the publicationApplied CAL to rapidly identify evidence for infection-prevention guidance during the COVID-19 pandemic.
Read about the review