AI's New Gold Rush: Unlocking Decades of Video Trapped on Tape

SAINT JOHN, New Brunswick and EAST PERTH, Western Australia – April 14, 2026 – In archives and vaults around the world, countless shelves are lined with aging video tapes, a vast and silent library of our visual history. For decades, this footage has been a depreciating asset and a growing cost. Now, a strategic partnership is set to transform these dormant archives into a foundational resource for the next generation of artificial intelligence.

Versos AI, a company building a data marketplace for the AI economy, has joined forces with Tape Ark, a global leader in mass media digitization. Together, they have announced a plan to create a direct pipeline from physical tape to AI-ready training data, potentially unlocking one of the largest untapped sources of real-world information for an industry with an insatiable appetite for content.

The Insatiable Appetite for Video Data

The AI boom, largely powered by text and images from the public internet, is entering a new phase. The most advanced systems, dubbed 'world models,' are being designed to understand and predict complex real-world dynamics, a task that requires an almost infinite amount of video. This is the critical bottleneck the new partnership aims to solve.

“The future of AI training data is video, and much of it is still trapped on tape and drives,” said Chris Keevill, CEO of Versos AI, in a statement. “This partnership with Tape Ark creates a direct pipeline to unlock that footage, and transform it into structured, usable data for AI systems.”

While current AI models have consumed much of the available online data, vast repositories of high-quality video remain offline, locked away by broadcasters, film studios, sports leagues, and major corporations. This content—from historic newsreels and unedited film takes to decades of sporting events—represents a trove of diverse, real-world scenarios that is far richer than much of the scripted or curated content available online. The demand is immense; the digital video content market is projected to surge past $500 billion by 2031, driven by AI-powered recommendation engines and the proliferation of streaming services.

By providing access to this previously unavailable data, the collaboration could significantly accelerate the development of more sophisticated AI capable of nuanced video understanding, generative video creation, and complex activity recognition for applications ranging from autonomous systems to medical diagnostics.

From Cost Center to Profit Center

For content owners, the economics of legacy archives have long been a one-way street of expense. Storing physical media in climate-controlled environments is a significant operational cost, with industry estimates cited by the companies putting the price between $0.40 and $1.20 per tape, per month. For organizations holding millions of tapes, this translates into a multi-million-dollar annual liability for assets that are practically inaccessible and degrading over time.

“Many are spending millions of dollars holding on to archive data that’s largely inaccessible, practically unusable and degrading every day,” noted Guy Holmes, CEO of Tape Ark. The partnership presents a radical shift in this paradigm. Instead of paying to store dormant tapes, content owners can now monetize them.

The process begins with Tape Ark, which has already processed over an exabyte of data globally, including 200 petabytes of video—equivalent to roughly 29 million hours of high-quality 4K footage. The company manages the complex logistics of physically restoring and digitizing a wide array of obsolete tape formats at an industrial scale, migrating the raw content to the cloud. This step alone slashes storage costs to a fraction of a cent per tape and makes the content accessible.

Once digitized, Versos AI applies its proprietary artificial intelligence to analyze the footage. Its platform automatically indexes video at the frame level, identifying and tagging objects, scenes, people, and events. A library that hasn't been viewed in years and was only cataloged with vague labels becomes a minutely detailed and instantly searchable database. This structured data is then made available for licensing on the Versos AI marketplace, allowing AI developers to acquire high-quality, relevant data for model training. This transforms a decaying cost center into a new, potentially lucrative, revenue-generating business.

Digital Archeology for the 21st Century

The initiative can be seen as a form of digital archeology, rescuing a significant portion of humanity’s visual record from the brink of oblivion. Many tape formats, like HDCAM-SR, have a functional lifespan of only 10-15 years, and the hardware required for playback is increasingly rare and difficult to maintain. Without intervention, this content risks being lost forever to bit rot and technological obsolescence.

The technical workflow established by the two companies creates a robust preservation and enhancement pipeline. Tape Ark’s mass migration to the cloud ensures the long-term integrity of the master files, while Versos AI’s enrichment process unlocks their hidden value. Archives that were once black boxes become transparent, analyzable assets.

However, this new frontier is not without its challenges. The quality of the source material can vary dramatically, and the digitization process must balance file size with preservation standards to avoid irreversible data loss. More significantly, content owners must navigate a complex web of legal and intellectual property rights. Determining copyright ownership, clearing rights for all individuals appearing in historical footage, and drafting licensing agreements specifically for AI training are substantial hurdles. The legal framework around using copyrighted material for training AI is still evolving, adding another layer of complexity for organizations eager to tap into this new market.

Despite these challenges, the joint effort by Versos AI and Tape Ark marks a pivotal moment. It establishes a new, vital data pipeline for the AI economy, connecting a wealth of previously inaccessible offline content to the systems that need it most. In doing so, it not only promises to accelerate the future of artificial intelligence but also provides a lifeline to preserve and monetize our collective visual past.