Understand the HISE Data Retention, Archival, and Deletion Policy—DRAFT
Last updated 2026-07-26At a Glance
To support applicable research record retention requirements described in its data governance policy, HISE retains data for a minimum of 6 years. HISE currently doesn't define a maximum retention period and doesn't follow a routine for deleting archived data. Future governance work might introduce a deletion process for archived data only.
Files are automatically archived when they haven't been accessed for at least 3 years, or when they have been marked for deletion or archival through supported workflows. In HISE, deleting data usually means soft deletion, not immediate permanent removal of the underlying file.
This data retention policy is part of the broader data governance framework that aligns with relevant NIST cybersecurity and data protection guidance used at the Allen Institute. If you have questions or need help, contact Support.
Scope
This policy applies to HISE-managed files and certain generated assets governed through ledger and hydration services. Covered data includes standard research files and some deleted draft assets created through IDE, visualization, remote job, and training job workflows.
Archival Process
Google Cloud Archive is the platform’s cold storage tier. Files automatically move to cold storage after about 3 years of inactivity or when marked for deletion or archival. Regular files are automatically archived after about 3 years of inactivity, based on last accessed date or last modification date. Some soft-deleted or low-value assets follow a shorter archival path of about 90 days, depending on availability and object type.
Files may also become archive candidates when they are explicitly marked for deletion or archival in supported workflows. Archived files remain searchable in some contexts, but they are not directly accessible. If archived data needs to be read again, an administrator must unarchive it and restore its pre-archival availability state. You can find archived files using Advanced Search, and then request unarchiving by contacting Support with the file IDs and project context.
NOTE
Archived storage is inexpensive for long-term retention, but access is expensive. Archived data also can't be deleted penalty free for at least 1 year, which is why archived data should be restored only when needed.
Deletion Process
In most workflows, deleting something in HISE doesn't immediately remove the underlying file from storage. Instead, HISE updates the item's availability state so it’s hidden from standard views, standard search results, or both, and the file is later archived by background routines. In the Project Store, you can mark files for deletion. Those files then disappear from standard views and remain in a “pending deletion” state for the next 30 days, during which time an administrator can still recover them if needed.
Depending on the workflow, an item may be marked with availability values such as markedfordeletion, markedforarchival, deleted, duplicatedeletion, superseded, archived, or archivedcruft. Not every value is user-settable in every workflow.