With recent advances in artificial intelligence, organizations increasingly recognize data as a vital resource for advancing scientific, economic, and societal progress. Many modern machine learning systems rely on large and complex datasets to learn patterns, generate insights, create new content, and support intelligent decision-making. However, the data needed to develop these models is often scattered across different organizations, with each holding only a portion of the necessary information. By sharing data with each other, instead of relying only on their own data, organizations can unlock new opportunities for discovery and innovation. Despite this potential, organizations often hesitate to share data unless there are clear incentives to do so. This raises important questions: Is it always beneficial for an organization to share data? Do participants receive benefits that reflect their contributions? What mechanisms can encourage data sharing while preventing participants from benefiting without contributing? When several parties collaborate to train a shared model, how should the benefits be allocated? This project addresses these questions by investigating the incentives that influence participation in data sharing and collaborative machine learning. The project aims to create environments where data sharing supports collaboration and innovation, ultimately enabling breakthroughs in areas such as health care, public policy, and education. This project will dev