Matthew Lease

Matthew Lease is a Professor at the School of Information, University of Texas at Austin, where he serves as Director of Doctoral Studies and Assistant Graduate Advisor. He is a Distinguished Member of the Association for Computing Machinery (ACM), a Senior Member of the Association for the Advancement of Artificial Intelligence (AAAI), and an Amazon Scholar. Lease co-directs the $20M NSF-Simons AI Institute for Cosmic Origins (CosmicAI) and is a faculty founder and leader of UT's Good Systems, an eight-year, $20M university-wide Grand Challenge aimed at designing responsible AI technologies. In 2023-2024, he was invited four times to address the Texas Legislature on responsible AI.

• Ph.D. Computer Science, Brown University, 2010
• M.Sc. Computer Science, Brown University, 2004
• B.Sc. Computer Science, University of Washington, 1999

Lease directs the UT Austin Laboratory for Artificial Intelligence and Human-Centered Computing (AI&HCC), where his research spans artificial intelligence modeling and human-computer interaction design. His work focuses on creating novel datasets, building AI models, and evaluating both model performance and their impact on end-users. When automated AI falls short, his team designs human-in-the-loop approaches, leveraging AI model explanations and creative user interfaces. To promote fair AI, they focus on better annotation techniques to avoid bias and develop modeling strategies to mitigate dataset biases. Their work tackles real-world problems as part of UT Austin's Good Systems Grand Challenge, with an ongoing emphasis on content moderation—exploring automated, human-in-the-loop, and human-safe practices to combat disinformation, hate speech, and online polarization.

Lease's recent publications (2021-2024) demonstrate a strong focus on human-centered AI, particularly in the areas of fair and explainable AI, content moderation, fact-checking, and crowdsourcing. His research integrates technical AI development with human factors considerations, emphasizing the importance of designing AI systems that work effectively with human users. The publications reveal a consistent theme of addressing bias in AI systems, improving human-AI collaboration, and developing methods to ensure the ethical deployment of AI technologies in sensitive domains like content moderation and misinformation detection. His work shows progression from foundational techniques in crowdsourcing and human computation toward more sophisticated approaches that consider psychological impacts and ethical implications.

• 2024 Test of Time Paper Award, AAAI Conference on Human Computation and Crowdsourcing (HCOMP)
• 2024 Most Influential Paper Award, IEEE/ACM International Conference on Automated Software Engineering (ASE)
• 2024 Best Paper Honorable Mention, ACM Conference on Computer Supported Cooperative Work (CSCW)
• 2022 Best Student Paper, Conference on Information Systems and Technology (CIST)
• 2020 Conference Award Track, Journal of Artificial Intelligence Research (JAIR)
• 2019 Best Student Paper, European Conference for Information Retrieval (ECIR)
• Early Career awards from DARPA, NSF, and IMLS

Lease has secured significant funding for his research, including the $20M NSF-Simons AI Institute for Cosmic Origins and the $20M Good Systems Grand Challenge. His lab, AI&HCC, has developed numerous tools and methodologies for human-AI collaboration, particularly in the context of content moderation and fact-checking. He has advised numerous students who have gone on to publish in top-tier conferences and journals in AI, HCI, and NLP. Lease actively collaborates with industry partners including Amazon, where he serves as an Amazon Scholar, and has served on advisory boards for JASIS&T, Texas Advanced Computing Center (TACC), and UT Austin-Amazon Science Hub. His research has led to practical tools like SQUARE for aggregating crowd responses and methods for transparent AI evaluation.

Lease leads the UT Austin Laboratory for Artificial Intelligence and Human-Centered Computing (AI&HCC), which has developed innovative approaches to human-AI collaboration. The lab's work on content moderation addresses critical challenges in online safety, including the psychological well-being of content moderators who face traumatic material. Their research on fair and explainable AI has produced methods for detecting toxic speech while maintaining accuracy across demographic groups. The lab actively collaborates with fact-checking organizations through co-design processes to create tools that meet real-world needs. As part of UT Austin's Good Systems initiative, the lab is developing AI technologies that prioritize human values and social responsibility from the outset of the design process.