What happened
In a press release dated October 28, 2026, Yandex said its engineers had restored access to 96% of data stored on Yandex Disk. The company says this total covers documents, photographs, audio recordings, video recordings, and other files. It presents the number as an overall service measure, not an account-level guarantee.
Access remains uneven among users. According to Yandex, some users have regained access to all of their files, while others have not. The company says it is returning files gradually and that more files will become available each day. It does not state how it calculates the 96% figure, whether it is measured by file count, volume, accounts, or another method.
Yandex says it is distributing capacity across independent third-party sites in order to maintain service stability during the recovery. The release does not name those sites or describe the storage, replication, routing, or restoration mechanisms involved. It also does not specify whether all files are available from the same locations as before. The stated operational result is staged restoration rather than a claim of complete recovery.
The company's listed core Disk functions are working: upload, download, and synchronization. But the browser version still has material limitations. Users cannot download folders or albums there, and they cannot download multiple files at once. Yandex directs users needing those operations to the Disk desktop application or the API. The announcement does not identify an API version or endpoint, announce a new API capability, or describe any change in API access terms.
Yandex Disk is part of the Yandex 360 virtual office. Yandex says other Yandex 360 services are operating with their main functions available, while teams continue to restore other capabilities gradually. It directs users to the Yandex 360 blog for progress updates. The press release gives neither a completion date nor a new price or access policy. It also does not state the cause of the disruption.
What the archive adds
This is a new recovery fact in a continuing service-continuity story. Yandex Cloud arranged accelerated placement of data and services for affected customers described a cross-provider recovery path. This announcement instead quantifies restored access to Yandex Disk files and lists remaining client limitations.
An earlier report on physical disruption at Yandex data centers is relevant context for the broader continuity story. The present Yandex update itself does not identify a cause, so it cannot establish a direct causal connection.
Trend signal
No supplied site pattern genuinely fits this narrowly scoped recovery update. The archive does reinforce a service-continuity concern: the earlier data-center report framed physical infrastructure disruption as a risk when services are concentrated, while this release shows access being restored progressively across independent sites. That is evidence of a recovery response, not proof of the underlying architecture or of future resilience.
Why now
The October 28 update is a progress report while recovery is ongoing. Yandex now provides a quantified access milestone, says which basic functions are available, and identifies browser operations that remain unavailable. It offers no schedule for complete restoration, so the practical significance is current status rather than a declared end to the incident.
Connections to other events
A prior Field Note reported that Yandex Cloud had arranged accelerated placement of data and services for affected customers. Taken together, the items describe different recovery tracks: placement and service continuity for cloud customers, and gradual return of files and functions for Disk users. Neither source says whether the tracks share infrastructure, vendors, or operating procedures.
Why it matters
Teams using cloud storage in an AI workflow should distinguish provider-wide availability from workflow-level usability. A storage service may support uploads, downloads, and synchronization while a browser client cannot perform the bulk retrieval operation needed to restore a project, move an evaluation set, or collect generated artifacts. Likewise, a 96% service-wide figure cannot establish that a specific account has its required files.
The concrete action is to test the paths a team actually uses. Where critical inputs or outputs sit in Yandex Disk, check access to the necessary folders and verify whether the browser, desktop application, or API supports the needed operation. Yandex's guidance toward the desktop application or API is a workaround for named browser limitations, not evidence that every automation or integration will behave identically. Independent copies of irreplaceable inputs and artifacts remain important when restoration is staged.
Limits and uncertainty
This is a company status update rather than an independent audit. Yandex does not define the denominator behind its 96% figure, say how many users remain affected, or provide a date for full recovery. It does not identify the cause of the disruption, the independent sites receiving capacity, or the architecture used to restore availability.
The release also does not establish that every Disk function has returned. It explicitly identifies unavailable browser bulk-download actions, and describes other Yandex 360 capabilities as returning gradually. It does not supply performance metrics, durability guarantees, service-level commitments, or account-specific status information. Nothing in the statement establishes whether an individual team's files, permissions, integrations, or automated workflows are fully operational.
Skill to build now
Make a short recovery drill for storage-backed AI work this week. Pick one critical project folder, an evaluation dataset, and a generated-artifact archive; confirm that each can be retrieved through the path your team would use during an outage. Then test an alternate route, such as a desktop client or approved API workflow, and record where an independent copy is held. The goal is not to assume that a provider-wide status number maps to your account, but to know which retrieval path works for the assets that matter.
Reader takeaway
Treat Yandex's 96% figure as a meaningful service-wide recovery milestone, not a promise that every account or workflow is complete. Core Disk operations are back, but browser bulk downloads remain restricted and restoration continues. Teams should verify their own critical files and retrieval routes before relying on the service as fully restored.
Open question
When will Yandex restore the remaining files and browser bulk-download functions, and how does it define the 96% access figure? The current release does not say.