A data trust that replaces the trusted intermediary with code and collective governance: the providers keep control of their data, smart contracts execute the agreements, and tokenisation carries the funding.
A classical data trust puts a fiduciary between the people who hold data and the people who want to use it. That works exactly as far as the trustee is trustworthy, and it concentrates the very power a data commons exists to distribute. The decentralised version keeps the fiduciary duty and removes the single point of trust: governance sits with the providers, typically as a DAO, and the agreements execute as smart contracts rather than as promises.
Three traits define it. Decentralised: data providers retain substantial autonomy and decide collectively. Trust-minimised: the contract executes transparently, so counterparty risk drops out of the arrangement instead of being underwritten by a custodian. Sustainable: tokenisation opens funding and value-sharing models that a centralised custodian cannot offer. It is the constitutional layer under Data Collaboration, and the mechanism that keeps the Research Loop on the right side of the two faces of increasing returns.
Source: Decentralized Data Trusts: A Trust-Minimized and Sustainable Approach to Data Collaboration, Rocketstar Foundation, April 2025.