Copyright law and intellectual property rights—the concept of automatically providing the creator of original work with full, exclusive, and inalienable rights to it that cannot legally be exploited—has been a major driver of innovation, development, and progress since the dawn of nations.
Today, however, its viability hangs in the balance. Copyright law is in desperate need of reform, and soon. Unfortunately, it is not being implemented quickly or well enough.
As mass internet connectivity and everything that has been built alongside it have shaped our modern world, copyright law has remained largely the same. Whilst we may be making great and unprecedented progress as our world becomes increasingly digital and governed by technology, certain technological advances threaten innovation, and there is one that is doing this perhaps more than any other—augmented (and by extension, artificial) intelligence.
Falling Foul of Copyright Law
Before we turn to why it is augmented intelligence that is causing one such a huge problem, it is a good idea to demonstrate just how easy it is to violate copyright law in a digital world where everything is social and capable of being shared.
Image courtesy of Human Resource Management Institute.
Examples of this are all over social media.
In a world where likes and shares can generate income*, content creators are no stranger to copyright infringement.
Facebook and Instagram pages are prime examples of this, and the so-called “meme pages” are by far the biggest culprit. When the administrators of these pages see something funny online–anything from a tweet to an artistic creation–all it takes is two minutes to take a screenshot of it and re-upload it to their own pages to benefit from the likes, shares, and comments.
*Whilst the page owner does not gain any financial reward through shared works directly, businesses can and have been built on the back of so-called social media "meme pages" through ad revenues, advertising deals, and merchandising. The content that the page owner shares is the main reason why individuals using social media platforms may be inclined to "like" and "share" the page and its content. More content posted equals more "likes". More "likes" equals more "shares" and more "shares" means more people will follow the page. This has a snowball effect and, before you know it, hundreds of thousands if not millions of people follow a page. It then becomes a cash cow.
There are several key questions raised here. Has a page's shared content—some of which violates copyright—contributed to its ability to make money? If so, by how much? Do rights holders have a claim to a proportion of this money? How do we quantify it?
This scenario and these questions are for another time, however, they demonstrate one key way copyright law falls short in our social and digital world.
Machines as Creators: The Problem with Augmented Intelligence
"There's a program for that" may not be a common phrase as of right now but, looking at the way things are going, it may well be soon. As major technological advances and innovations continue to fly at us at warp speed with each passing day, more and more programs, systems, and machines are being pioneered that are not conquerors, but creators.
Image courtesy of Autodesk.
Computers that can analyse artwork and then generate its own, artificial intelligence programs that can write their own articles, and even generative design tools that make the jobs of engineers easier by creating hundreds to thousands of different product design iterations when fed with parameters are all challenging copyright law.
These three are but a few of many examples of the role computers are increasingly beginning to play—they are no longer tools that merely facilitate the work of human beings in the way that Adobe Photoshop does, they are creators of original works in their own right… but a machine cannot possess the right to works, so the question is thus—who does?
Copyright Ownership in EU Law
Rewind a decade or two and the question of who owns copyright to creative works involving a computer was not a difficult question—the human creator did; it was his or her input using the computer as a tool to merely aid the creation, not create it completely.
As we have just seen, this is no longer the case. Let's look at what the law says.
- Under UK Law—specifically the Copyright, Designs and Patents Act 1988—it is stated: "In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken"—Section 9(3).
- Under EU Legislation—Article 2 of the Copyright Directive—the author is created, and all Member States must exclusively provide rights to prohibit the infringement of creative works in whole or in part, temporarily or permanently, directly or indirectly.
- Under EU Case Law—Infopaq International A/S v Danske Dagblades Forening—the author has been held to only include human creators through the court's decision to apply copyright only to original work where said originality reflects the "author's own intellectual creation". An author's own intellectual creation is partly comprised of their personality, something a computer, system, program, or machine cannot have.
Unfortunately, none of this helps us answer the question of who should rightfully own copyright where AI is used to create works. Legally, of course, nobody does—it is in the public domain.
The Spanner in the Works
There is a clear need for the EU to act and revise its copyright law.
Excluding non-human authors from being protected by copyright could stifle progress and innovation.
Companies are already using augmented and artificial intelligence in the production of their works, however, under the current legal framework, the creators of these augmented and artificial intelligence computers, tools, and systems have no title to the very works that they have, in part, created. Under the author principle of EU law, these AI-created works are essentially free of copyright entirely—not even the company using the artificial and augmented intelligence solutions to create a work have rights to it.
Without protection, even temporary protection that facilitates a temporary monopoly over works, the creators of artificial and augmented intelligence have little incentive to continue developing or improving upon their creations; the lack of financial benefit will surely limit innovation across all industries.
What is the Optimal Solution?
Whilst it has been suggested that the best solution for this problem is to take the UK's law as is—where the author of a work is the person who made the arrangements necessary for its creation, e.g. by using, configuring or 'programming' a tool to create a piece of work—for computer programming-based artistic works, the potential problem of stifled developments in AI cannot be ignored.
If person A spends months, if not years, developing a tool that person B then freely uses to create and make money with an artistic work, how is it fair that person A holds no copyright, not even temporarily, and is not entitled to some remuneration?
It simply isn't. After all, without person A's tool, person B's work would not exist, at least not in the same form.
Image courtesy of Electronic Products.
If we are to benefit from further strides in augmented and artificial intelligence, legislation needs to be revised that recognizes the rights of both parties. People will not work for free and as more and more developers begin to focus on creating powerful AI solutions that are able to create works in their entirety, it will reach a point where advances in these and other tools will stop dead in their tracks.
Comprehensive legislation that changes the EU's current position of non-human authors is needed to prevent stagnation of AI development. It will not be easy, though, and many questions will be born from ambiguity.
Returning to the example above where a computer was able to analyse and re-create a painting, who can be attributed to its copyright? The creator of the original painting? The programmer? The person who set the program's parameters?
Whilst this article cannot analyse this issue in the depth it warrants and it by no means comes close to even touching on a solution—a task fit for legal minds far greater than mine—what is clear is that AI has not even left the starting gate yet but it will very soon, very suddenly, and new legislation and EU legal framework for copyright is needed sooner rather than later.