Court Rules that AI Art Can't Be Copyrighted

You’ve probably seen them clogging up your social media feed and any other highly-trafficked visual platforms – AI art, or artistic works generated via artificial intelligence. And now, for the very first time, a federal judge has issued a ruling on copyright protections for AI-generated art (spoiler alert – it isn’t covered).

In his ruling, Judge Beryl Howell of the US District Court for the District of Columbia noted that “Courts have uniformly declined to recognize copyright in works created absent any human involvement.”

As it stands, and according to the U.S. Copyright Office, the pictorial, graphic, and sculptural works eligible for copyright include “two-dimensional and three-dimensional works of fine, graphic, and applied art, photographs, prints and art reproductions, maps, globes, charts, diagrams, models, and technical drawings, including architectural plans.”

But there’s always been the assumption of “human authorship,” effectively keeping all AI art in the public domain. This latest ruling reinforces that.

What does this mean? Well, there’s the obvious – you can’t copyright AI art, but on the flipside, it also makes corporations less likely to pay human artists. Why pay someone when they can scoop up AI art (or generate their own) from across the web, with no consideration of a non-existent copyright?

We’re not quite at the point where all AI art is indistinguishable from the real thing. Not yet.

Back when Alan Turing originally proposed his “imitation game” (aka, the Turing Test), where a machine would try to pass as a human, most of us never figured that AI would become this…intelligent.

At the least, the creative arts seemed to be beyond the purview of machines.

But no longer – not only can ChatBots reliably pass Turing’s test, but we’ve reached the point where poetry, essays, paintings – anything creative, really – seems attainable for artificial intelligence.

As mentioned, we’re not there yet. There’s something subtly off about AI-generated human portraits, whether it’s oddly-shaped hands or weird proportions. An entirely AI-generated beer commercial was making the rounds recently, and it looks suitably horrifying, like Skynet’s idea of a David Lynch movie.

But as AI absorbs more information – ironically, learning from copyrighted works – its creativity becomes more and more pronounced, and quite simply, more human.

Something called AICAN (AI Creative Adversarial Network) was apparently trained on 100,000 images over 500 years of Western art, and its original creations were indistinguishable from expressionist masters and contemporary art.

If we’ve truly reached the point where AI can replicate feasible art, and if said art can’t be copyrighted, the entire artistic industry could soon turn on its head.