Nine PBS Data Loss: What Happened to 50 TB of TV History?

National television history is only part of the story here, and arguably the smaller part. A local public television archive built over 70 years contains things no national network ever recorded: neighborhood meetings, regional elections, local schools, community organizations, and the kind of on-the-ground coverage of events like the 1993 Great Flood or East St. Louis that never aired outside the region. People who never appeared on national television — local officials, teachers, activists, ordinary residents interviewed for a community-affairs program — exist on tape (and later, on file) largely because a station like Nine PBS chose to point a camera at them.

That is the real stake in this story. The strongest question isn’t “is cloud storage dangerous?” — it’s what happens when a community’s memory becomes dependent on a company that no longer controls the storage relationship. A television station entrusted decades of local history to a third-party storage arrangement, and when that vendor relationship broke down, access to a huge archive suddenly became uncertain. That’s a story about institutional memory, vendor dependency and data ownership — not a referendum on cloud computing as a technology.

Two companies sit at the center of this story, and they are not the same thing. Open Source Storage (OSS) was Nine PBS’s contracted vendor — the company the station paid, since 2019, to store its digitized archive. Iron Mountain is a large, established data-management and storage company that, per reporting, physically hosted OSS’s own servers inside one of its Denver data centers. Nine PBS had no direct contract with Iron Mountain at all — its relationship was entirely with OSS.

The available reporting does not establish a single, confirmed reason for OSS’s collapse. What is reported: OSS stopped responding to Nine PBS in February 2026, its public website went dark, and Colorado state business records reportedly list the company as delinquent. Its president is identified in reporting as Charles Wells. Whether this reflects a bankruptcy, an orderly wind-down, an internal dispute, or something else has not been established by available reporting, and this article does not speculate beyond what has been reported.

Iron Mountain’s position, per its own public statement, is that it dealt appropriately with a difficult situation: its contract was with OSS, not Nine PBS, and OSS — not Iron Mountain — technically owned the physical servers housing the station’s data within Iron Mountain’s facility. That contractual structure is precisely why a station whose vendor goes dark can end up with no one to call: the party holding the hardware has no direct obligation to the party that actually owns the content on it.

“Our archive is in the cloud” is not the same statement as “we have a recoverable backup.” Primary storage is wherever the working copy of a file lives day to day. A backup is an independent copy, ideally on different infrastructure, that survives the loss of the primary copy. Replication mirrors data across locations for uptime, not necessarily for independent recovery. An archive is long-term storage of material not in active use — which is what Nine PBS’s collection was. Cold storage and offline archives add a further layer of separation, often intentionally hard to reach quickly, in exchange for durability and lower cost.

Vendor-controlled cloud storage can be any or none of these, depending entirely on the contract and the architecture behind it — and that’s the trap. When a single vendor holds the only accessible copy of an archive, that vendor relationship is the recovery plan, whether or not anyone framed it that way at signing. Nine PBS’s files were not, per current reporting, deleted or corrupted; the problem was access — the contractual and infrastructural relationship that made the files reachable stopped functioning. That is a distinct failure mode from data loss in the literal sense, and it is arguably a more common one for organizations that outsource archival storage.

Data can remain fully, physically present on a disk somewhere and still be completely unreachable to the organization that owns it. That’s the situation reporting describes for Nine PBS: the archive appears to still exist inside Iron Mountain’s Denver facility, on servers that belonged to OSS. What Nine PBS lost first was not the bits — it was the login credentials, the account relationship, and the contractual path to the hardware. A vendor’s infrastructure can change, its authentication systems can be shut off, its billing and account systems can go dark, and a customer’s catalog or metadata system can become unreachable even while the underlying files sit untouched a few feet away in a rack.

Digitizing an archive solves some preservation problems — tape degradation, obsolete playback hardware, physical damage — and creates new ones. File formats age; a video codec common a decade ago may need conversion to stay usable. Metadata — the catalog information that says what a file actually is, when it was recorded, and why it matters — is as important as the video itself, and is easy to lose or disconnect from the material it describes. Storage costs money indefinitely, not once; “bit rot” (silent data corruption over time) requires active checksum verification to catch; software dependencies for a rights-management or catalog system can break when that software is discontinued; and true resilience requires multiple copies in geographically separate locations, actively maintained, not just initially uploaded and forgotten.

None of this is unique to Nine PBS. It is the standard, well-documented challenge facing every institution — broadcaster, library, museum, government archive — that has moved from physical to digital preservation. What the Nine PBS case adds is a sharp, concrete illustration of one specific failure mode within that larger challenge: outsourcing the custody relationship, not just the storage medium.

This article is based on independent reporting from technology and public-broadcasting trade press, cross-checked across multiple outlets. It does not rely on Reddit or other unverified discussion forums for factual claims — those are used, if at all, only to characterize public reaction, never as a source of fact.