Public data licensing operation is an important component of the planning for the development and utilization of public data resources. However, there are still deficiencies in the current licensing operation in terms of responsibility allocation model, data ownership distribution scheme, and the smoothness of the circulation mechanism. The improvement approach should be to construct a system framework that combines theoretical explanatory power with practical adaptability. Data trust relies on its dual attributes of legal framework + technical governance. Through empirical tests of projects such as the UK’s ODI and the US’s One Florida+, it demonstrates innovative value in addressing the aforementioned dilemmas. In terms of licensing forms, compared to general entrustment, trust better reflects the purpose and pursuit of public data licensing operation: the management and long-term characteristics of trust relationships align with the cyclical pattern of data from generation to flow to processing and utilization, and also show institutional potential that matches licensing operation activities. In terms of the nature of the object, data trust has interpretability regarding the ownership of public data: both trust property rights theory and data property rights separation belong to the derivative of the bundle of rights theory, and the trust system, as a path for allocating data-related rights, has inherent adaptability. In terms of the feasibility of legal interpretation, firstly, public trust theory embodies the orientation of actively pursuing and passively safeguarding the value of public property. Public data with public property characteristics also have the necessity and feasibility of being managed in the form of trust. Secondly, the certainty of trust property is a legal prerequisite for the establishment of trust. The data registration system can ensure the independence of the data portion involved in trust through inventory management and technical traceability, clarify the correspondence of ownership with human compilation doctrine, and, in collaboration with the existing trust registration system, utilize a centralized trust registration institution to provide publicity and effectiveness support for the trustification of public data. Thus, the scope of public data invested in trust can be determined, thereby meeting the necessary conditions for the circulation of trustified public data. In terms of specific relational construction, further concretization and standardization of various relational elements in trust construction are needed: the trustor should enjoy the right to supervise that is differentiated from administrative supervision, and undertake the obligation to supply public data with reference to the separation of collection and transportation model; the trustee should enjoy the right to profit from the use of public data and undertake the fiduciary duty to fulfill the trust contract and maximize the beneficiary’s interests. In terms of trust income distribution, it is necessary to reasonably confirm the allocation scheme of operating income among public entities, specific data sources, and unspecified members of the public, achieving a balance between public welfare and profitability. In terms of liability determination, it is advisable to distinguish between the damage to the rights and interests of third parties caused during operations and the damage liability within the trust relationship, and to refine the determination of liability for each subject in different scenarios.