I’ve been trying to track down what the heck happened in 1965 in copyright & computers that the Ars Technica piece I referenced earlier places such importance on vis a vis the USCO & computers in art. The Ars piece doesn’t actually say what it was that happened, just this:

But the Copyright Office insisted that the AI copyright debate was settled in 1965 after commercial computer technology started advancing quickly and “difficult questions of authorship” were first raised. That was the first time officials had to ponder how much involvement human creators had in works created using computers.

Skimming through the actual report by the USCO yields not a lot more detail:

Given its role in registering claims to copyright,5 the Copyright Office has considerable experience addressing technological developments related to the creation of works of authorship. As early as 1965, developments in computer technology began to raise “difficult questions of authorship,” including whether material created using technology is “‘written’ by computers” or authored by human creators.6 As then-Register of Copyrights Abraham Kaminstein observed, there is no one-size-fits-all answer:

“The crucial question appears to be whether the “work” is basically one of human authorship, with the computer merely being an assisting instrument, or whether the traditional elements of authorship in the work (literary, artistic, or musical expression or elements of selection, arrangement, etc.) were actually conceived and executed not by man but by a machine.”

This just sounds like more of the “it depends” style of rule-making, which seems still not adequate to me, whatever the year.

Thanks to a ChatGPT conversation though, I did manage to trace down this interesting 2016 paper by Annemarie Birdy, entitled “The Evolution of Authorship: Work Made by Code” (PDF linked at bottom of that page). Some cobbled together quotes from that:

In 1965, Register of Copyrights Abraham Kaminstein identified the question of computer authorship as one of three major problems confronting the Copyright Office. A number of people filed copyright registrations that year for works at least partly authored by computers, and the Office found itself at a loss for how to deal with the situation.

Later:

But does copyright law require human creativity? The Copyright Act doesn’t say anywhere that an author has to human, and there’s really no case law directly on point. Nevertheless, there seems to be an assumption, maybe driven by practical and historical considerations, that authorship means human authorship. The 1965 Register of Copyrights Annual Report [here’s the document] frames the question precisely in terms of the human-computer divide. If a human creates a work, it’s copyrightable. If a machine creates it, then it’s not. The CONTU report does the same thing: unless there’s minimal human creative effort, there’s no protection.

Okay, so this is absolutely mind-blowing that in the actual Copyright Act, there is no requirement that authorship is human-only… The piece talks about a 1956 attempted copyright registration of computer-assisted song-writing:

The only problem that Klein and Bolitho encountered in
their digital composing project was that the Copyright Office refused registration for “Push Button Bertha,” one of Datatron’s many compositions. The reason the Office gave at the time was that no one had ever before tried to register a piece of music written by a machine. The rejection, for which the Office didn’t offer—and couldn’t have offered—any statutory basis, revealed a deep-seated if unspoken
assumption that authors are necessarily human.

Then in a footnote it says this assumption wasn’t codified until 1973:

By 1973, which brought publication of the first Compendium of U.S. Copyright Office Practices, that assumption had become explicit. See U.S. COPYRIGHT OFFICE, COMPENDIUM OF COPYRIGHT OFFICE PRACTICES (FIRST) § 2.8.3 (1st ed. 1973) (stating that works are not copyrightable if they do not “owe their origin to a human agent”).

    As a layperson, its difficult to navigate what official standing the COMPENDIUM OF PRACTICES must have relative to the actual law, but it must be… non-zero?

    According to ChatGPT – so this must be taken with heaping grains of salt – but it “sounds right” approximately:

    The Compendium of U.S. Copyright Office Practices is an administrative manual that outlines the practices and procedures of the U.S. Copyright Office. While it serves as a comprehensive guide for the Office’s staff and the public, it does not have the force of law. However, courts may consider it as persuasive authority due to the Copyright Office’s specialized expertise. 

    So…. that’s interesting?

    But going back to the Birdy piece, it follows on with an, I think, devastating point (and one which I included in my submission to the USCO originally), that the law already clearly accomodates for “non-human” authorship by way of work-for-hire arrangements, such that a corporation becomes considered legally as the author of the work.

    What the anthropocentric view of authorship elides, however, is that copyright law already accommodates a notion of non-human authors; they’re called corporations. Under the work-made-for-hire doctrine, which is a legal fiction, a corporate employer is considered the legal author of a work of which it is not the author-in-fact.38 The statute could have been written to create an assignment by operation of law from an employee-author to her corporate employer, thus maintaining in principle a human monopoly on authorship, but it wasn’t. It was written to allow a corporate employer to be treated ab initio as the author of a work created by its human employee.

    Because we already have a copyright doctrine that accommodates non-human authors, maybe that’s a logical place to look for a solution to the problem of computer authors. Maybe we can treat computer-authored works as works made for hire.

    So there we are. And that paper is from 2016. So these thoughts have been in the air for quite some time. But still we’re stuck on following a “practices” document, instead of explicitly following the actual law. I’d call that a pretty big discovery for me personally, as these things go…

    Related writing & research