Sustainable AI: New Paper Explores How Organizations Can Use AI More Sustainably
October 27, 2026
A new paper in the 2026 Joint Statistical Meetings (JSM) Proceedings examines how statisticians and data scientists can help organizations use artificial intelligence (AI) sustainably. Sustainable AI: Deployment and Opportunities for Statistical Practice highlights practical approaches for balancing AI innovation with responsible use, including model optimization, governance frameworks, design principles minimizing data movement across complex systems, and strategies for deliberate measurement to guide improvements in system efficiency. The authors emphasize the important role of statistical practitioners in furthering understanding of these complexities and guiding decisions that improve efficiency without compromising performance.
The paper arose from an invited panel organized by Westat’s Liz Mannshardt, PhD, Vice President and Director of Statistics and Data Science, and Elizabeth Eisenhauer, PhD, Senior Statistical Associate. Mannshardt moderated the panel, which brought together leaders from academia, industry, and government to discuss emerging research and practical approaches to sustainable AI practices. Panelists included Mahmut Kandemir, PhD (Penn State); Cynthia Rudin, PhD (Duke); Steve Sain, PhD (Jupiter Intelligence); and Lyndsay Shand, PhD (Sandia National Laboratories).
“AI has enormous potential to improve decision-making and accelerate innovation, but we also have a responsibility to understand and manage its resource demands,” said Mannshardt. “Statisticians are uniquely positioned to help organizations evaluate trade-offs, develop effective governance practices, and implement AI solutions that are both impactful and sustainable.”