Who Owns the Ghost? The Unresolved Crisis of Preserving Films Nobody Claims

Somewhere in the United States, a film is dying. Not dramatically, not with any particular announcement. It is simply deteriorating — acetate shrinking, magnetic oxide flaking, color shifting toward a terminal magenta cast — in a storage unit, a church basement, or a regional archive that lacks the budget to do anything about it. The film may have screened once, at a regional festival in 1974, or played a single midnight run at a now-demolished theater in Cleveland. No one has claimed it. No distributor holds its rights. Its director may be dead, or simply unreachable, or entirely unaware that anyone might still want to see it.

This is the orphan film problem, and it is not a niche archival concern. It is one of the central crises facing the preservation of American independent and underground cinema — and it is growing more complicated, not less, as new technologies introduce new possibilities and new legal ambiguities simultaneously.

The Scope of What Is Being Lost

The Library of Congress has estimated that approximately seventy percent of all silent films produced in the United States have been lost entirely. The survival rate for sound-era independent and underground films, while better, remains deeply troubling. Films produced outside the studio system — regional exploitation features, no-budget experimental works, community-funded documentaries, underground narratives shot on 16mm in urban lofts — were rarely preserved with institutional rigor. Their rights were poorly documented from the outset, and the passage of decades has turned poor documentation into genuine legal opacity.

An orphan film, in the technical sense used by archivists and legal scholars, is a work still nominally under copyright but whose rights holder cannot be identified or located. Under current American copyright law, this status creates a near-paralysis. An archive that digitizes and distributes an orphan film without clearing rights risks infringement liability even if the rights holder has been entirely absent for fifty years. The law, as written, protects the ghost of ownership rather than the work itself.

The practical consequence is that institutions with the resources to preserve these films are frequently unwilling to take on the legal risk of making them accessible. The films are stabilized, sometimes, against further deterioration — but they are not shared. They sit in climate-controlled limbo, technically saved and functionally invisible.

What AI Restoration Changes — and Complicates

The emergence of AI-assisted film restoration has introduced a new dimension to this already fraught landscape. Tools capable of algorithmically reconstructing damaged frames, stabilizing degraded footage, reducing grain and noise, and even extrapolating missing color information have matured significantly over the past several years. What once required the labor of skilled technicians working frame by frame can now be partially automated, reducing the cost of restoration to a point where it is theoretically accessible to small archives, independent organizations, and even dedicated individuals.

This is, in one sense, genuinely transformative. For the orphan film sitting in a deteriorating state with no institutional patron, the possibility of low-cost AI-assisted digitization represents a lifeline that did not previously exist.

But the legal complications introduced by AI restoration are considerable. When a machine learning model is applied to a damaged film — filling in missing frames, reconstructing degraded image data, generating approximations of lost visual information — the question of what, precisely, the resulting artifact represents becomes legally murky. Is the restored version a faithful reproduction of the original work, or a new derivative work? If it is derivative, who holds the rights to it? The restorer? The AI tool's developer? The original rights holder, if one can eventually be found?

These questions do not yet have settled answers in American intellectual property law, and the institutions best positioned to test them in court — major studios, well-funded archives — have little incentive to do so on behalf of obscure underground films.

The Internet as Accidental Archive

In the absence of institutional solutions, a preservation ecosystem has emerged from below. Online communities — operating across platforms ranging from dedicated film forums to video-sharing services to private Discord servers — have quietly become significant repositories of underground and independent cinema that formal institutions have declined to protect.

The ethics of this activity are genuinely complex. Much of what these communities preserve and share exists in legal gray zones. Films are uploaded without rights clearance, circulated without institutional sanction, and maintained by individuals whose archival credentials consist entirely of enthusiasm and hard drive space. From a strict intellectual property standpoint, significant portions of this activity constitute infringement.

And yet the historical record is unambiguous: a substantial number of underground and independent American films survive today because someone, somewhere, digitized a deteriorating tape and posted it online before the physical source was lost. The Library of Congress and the Academy Film Archive have both, on occasion, sourced restoration materials from online uploads when no institutional copy existed. The informal network has, in multiple documented cases, outperformed the formal one.

This creates an uncomfortable institutional posture. Archives that would never officially endorse unauthorized digitization are, in practice, benefiting from it. The preservation community is aware of this tension and largely declines to resolve it publicly.

Orphan Film Legislation and Its Limits

Efforts to create a formal legal framework for orphan film preservation in the United States have a long and frustrating history. Proposed orphan works legislation has been introduced in Congress multiple times over the past two decades, most recently in forms that would limit liability for good-faith users who conduct a diligent search for rights holders before using an unclaimed work. Each iteration has stalled, largely due to opposition from rights-holder advocacy groups concerned about the precedent such legislation would set for their own catalogs.

The result is a policy vacuum that disproportionately harms the most marginal works — precisely the underground and independent films whose cultural value is least legible to the commercial interests that shape copyright policy.

Some archives have pursued alternative approaches. The concept of "last resort" preservation — digitizing and providing access to orphan works on the grounds that the alternative is permanent loss — has been advanced by several academic and independent institutions as a defensible ethical position, even in the absence of legal clarity. The argument is straightforward: a copyright that protects a film into nonexistence serves no legitimate purpose.

The Keepers the Institutions Left Behind

What emerges from a clear-eyed survey of the current landscape is that the real preservation of forgotten American cinema is being carried out, in large measure, by people and communities that formal institutions have not empowered and, in some cases, actively discourage. Individual collectors, online archivists, small nonprofit organizations operating on minimal budgets, and dedicated communities of cinephiles are doing work that the Library of Congress and the major film archives cannot or will not do.

This is not a comfortable conclusion. Informal preservation is, by its nature, inconsistent, legally precarious, and dependent on the continued enthusiasm of individuals who may not be there in twenty years. It is not a substitute for institutional commitment and clear legal frameworks.

But it is what exists. And for the underground films that have no institutional patron, no living rights holder, and no commercial value to attract one, it may be the only preservation that ever happens.

The ghost deserves better than this. The question is whether American copyright law, and the institutions it governs, will ever be willing to say so.