AI Institute for Engaged Learning

The renewal of the NSF AI Institute for Engaged Learning (EngageAI) will leverage rapidly emerging artificial intelligence (AI) technologies to realize the transformative potential of AI for K-12 education. The Institute will create innovative AI technologies that promote effective, AI-enabled learning and teaching for in- and out-of-school settings and contribute to the development of the future science, technology, engineering, and mathematics (STEM) workforce. Building on the strong foundation of the Institute’s prior work on narrative-centered learning technologies, collaborative narrative-centered learning, and multimodal learning analytics, EngageAI will advance both the theory and practice of AI-enabled education, with a disciplinary education focus on science education and AI literacy. Recognizing that agency has a consequential role in driving human learning, the Institute will produce two families of advances in AI-enabled learning and teaching. First, self-adaptive learning environments that leverage AI world models will continually improve their capabilities to create story-based experiences tailored to the cognitive and social needs of students and teachers. Second, multimodal authoring tools will promote agency-centered teaching and learning by enabling both teachers and students to create learning environments via multiple modalities such as natural language, sketching, photos, and diagrams. In addition to its research and development agenda, EngageAI will serve as a nexus for AI-enabled narrative centered learning by bringing together people, ideas, problems, and technical approaches for maximum impact beyond the boundaries of the Institute itself. Central to these plans will be its community-building initiative, which will involve the League of Innovative Schools (over 150 school districts, nationwide), Association of Science and Technology Centers, OpenSciEd, SafeInsights, the EdSAFE Industry Alliance, and a network of public, private, and corporate foundations, all dedicated to the build-out and promotion of wide-scale adoption of these technologies in K-12 schools nationwide. EngageAI will produce transformative advances in AI-enabled agency-centered learning and teaching by bringing together a team of multidisciplinary researchers from North Carolina State University, Indiana University, University of North Carolina at Chapel Hill, Vanderbilt University, as well as a national educational innovation non-profit organization, Digital Promise. The Institute's use-inspired AI research will investigate, across studies in schools and museums, how AI world models can power: (1) self-adaptive narrative-centered learning technologies that improve students' science learning, AI literacy, and sense of agency; and (2) multimodal narrative-authoring tools that help teachers create engaging, classroom-tailored learning experiences while improving their own AI literacy and sense of agency. Complementing and interleaved with the Institute’s use-inspired AI research, its foundational AI research will produce significant advances in AI world models, which are emerging as the next frontier of AI. These will include foundational AI advances in continual learning, generative visual hypothesis testing via video diffusion world models, and long-term behavioral memory for world models. Woven throughout the Institute’s activities will be a strong focus on ethics, with an emphasis on creating AI-enabled learning that is deeply informed by considerations of fairness, accountability, transparency, trust, and privacy. Together, these efforts will drive institute-scale R&D to produce mature, narrative-centered learning technologies that leverage AI world models; a compelling body of evidence for their efficacy, technical and architectural advances; and national intellectual research leadership on AI-enabled agency-centered learning and teaching. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.