Zero-copy data sharing

Zero-copy data sharing lets one organization or team grant another direct query access to a dataset without duplicating it — no export file, no copy job, no second storage bill for the recipient. The data stays in the provider's storage, and the recipient's queries run against that same underlying data, typically governed by the same platform's access controls.

This differs from traditional data sharing, which moves a snapshot or extract to the recipient's own system through a file transfer or an ETL job, creating a second copy that immediately starts drifting out of sync with the source and that must be secured and refreshed separately. Zero-copy sharing instead relies on either a shared underlying storage layer, common on platforms built on an open table format, or metadata-level pointers that let a separate compute environment read the same files directly.

It matters for data-sharing relationships between organizations, business units, or a vendor and its customers, because the recipient always sees current data with no replication lag, and the provider avoids the cost and governance burden of managing exported copies; it is a foundational capability behind data mesh approaches to sharing domain-owned data. The main limitation is that it generally requires both parties to be on the same platform or a compatible storage and metadata layer, which is less flexible than a file export that any system can consume.