Microsoft is about to make a quiet change to Fabric that most tenants will never notice, but it will save a few of them from real pain. Starting August 23, 2026, Fabric will enable Item Recovery by default for any tenant that has not explicitly configured the setting. For supported item types, that means a default three-day window to bring deleted work back. It is the kind of default that only matters on the day you need it, and on that day it matters a lot.
What changes
Up to now, Item Recovery was something an admin had to turn on and decide a retention period for. Tenants that did nothing simply had no recovery window, which meant a deleted workspace or report was gone for good. The new default gives every tenant a base level of protection: three days of recovery for the supported item types, with no action required.
The important qualifier is “tenants that have not explicitly configured the setting.” The change is not forcing anyone. If your admin already set a retention period, that setting stays untouched. The default only fills the gap for tenants where nobody ever made a choice. Microsoft is closing a footgun rather than overriding a deliberate decision.
For tenants that want more breathing room, the recovery window is configurable well beyond the default. Admins can adjust retention in the Fabric admin portal, with options reaching out to a much longer range. A three-day default is the sensible minimum; a team that does large destructive refactors of workspaces may prefer a longer window.
Why the default matters
Most data loss in these platforms is not a case of attackers or hardware failure. It is a human deleting the wrong thing, or a script that removed a workspace it was not supposed to touch. A recovery default is cheap insurance against exactly those moments. The cost of keeping deleted items around for a few days is small. The cost of a permanently deleted dataset that took weeks to build is not.
The timing of the change is worth a note too. It lands after Item Recovery reached general availability earlier in the year, once Microsoft had confidence in the feature itself. Turning a reliable feature on by default is a reasonable call. The default is short enough not to be a burden, and long enough to cover the classic “I just deleted this by accident” scenario.
What admins should do
The immediate task is to check what your tenant currently has, because the change is about to make the previous behavior a thing of the past whether you acted or not. Open the Fabric admin portal and look at the Item Recovery setting. You will see three states: on with a configured window, off explicitly, or the unset state that is about to inherit the default.
From there, decide whether three days is right for your organization. Data governance teams often want to align the recovery window with their retention policy, and it is easier to set that up now than to learn the default is too short after something important is deleted. There is no downside to setting an explicit window that matches your needs, since an explicit choice is what overrides the default cleanly.
For the majority of smaller tenants, the honest answer is that this change requires nothing. You get three days of protection for free, and the admin portal will still show the setting if you ever want to extend it. The only wrong move is to ignore it entirely, because the day will come when a workspace vanishes and somebody asks why there was no safety net.
How this compares to other platforms
Fabric is not first to this pattern, and it is worth seeing the change in that light. Azure DevOps has long kept deleted work items and repos in a soft-delete state recoverable by admins, and Teams gives a default recovery window for deleted channels and chats before permanent removal. The cloud providers that ship these defaults tend to converge on the same reasoning: deleted state is cheap, permanent deletion is irreversible, and leaning toward recovery is almost always the safer bias. Fabric joining that group is overdue rather than novel.
The contrast is with systems that hold a hard line against soft delete for compliance reasons. Some orgs want deleted data gone immediately and cannot tolerate a recovery window that keeps it around, even briefly. That is precisely why Microsoft built this as a configurable, admin-controlled setting rather than a hard default. Teams with strict deletion requirements set the window to zero or disable recovery explicitly. Everyone else gets a safety net that costs nothing to hold.
The takeaway
Defaults quietly shape how much damage carelessness can do. Microsoft is nudging Fabric toward a state where a basic recovery window exists everywhere by default, which is the right instinct. Given how cheap it is to keep deleted items around for a few days, three days of protection is a low bar, and tying it to a configurable, admin-controlled setting keeps it from becoming a compliance problem. If you are a Fabric admin, spend two minutes reviewing the setting before the August date rolls around, and set an explicit window that your org is comfortable with.