Democratizing responsible artificial intelligence for innovation and impact

Artificial intelligence (AI) is driving discovery, innovation, and economic growth across most sectors and has the opportunity to transform society by addressing important grand challenges, such as predicting and managing natural disasters or disease outbreaks, and developing cures for diseases. While AI provides opportunities for major positive impact, it also has the potential to lead to negative social, environmental, and economic consequences if it is not pursued responsibly. To ensure the responsible progress of AI, specific consideration must be given to ensure the development, access, and impacts of AI are broad and inclusive. Such a democratization of AI is only possible through partnerships that bring together workforce development, research, and cyberinfrastructure resources, along with appropriate policies and regulations to ensure safe and ethical use. These partnerships must be supported by diverse teams of researchers with different perspectives and approaches, and through engagement with the broader public. In this paper, we explore the importance of democratizing AI research and innovation and the role of democratizing access to advanced cyberinfrastructure, including high-performance computing, in achieving this goal. Specifically, we introduce the University of Utah’s Responsible AI Initiative (One-U RAI) as a concrete example and discuss a recent effort aimed at achieving these goals through public-private partnerships across government, academia, and industry. The One-U RAI provides the framework for responsible, transdisciplinary AI development for societal impact in the areas of environment, health, and education, and is built upon a foundation of research expertise, advanced cyberinfrastructure, and community engagement to catalyze excellence in responsible AI.