Auctioning Art(ificial Intelligence): The IP implications of Edmond de Belamy

The IPKat has had his portrait painted by the
AmeriKat - raising another issue of
non-human authorship/ownership
In the AmeriKat's free time (which does, on occasion, exist), she takes her paw to some painting, being her medium of choice since kittenhood.  The paintbrush is dictated by the AmeriKat's paw which, in turn, is dictated by the spark of artistic expression and emotion lighting up the hemispheres of her brain.  Artistic expression which cannot be downloaded, decoded or analyzed in any complete way.  It is this artistic expression and the old image of the tortured, hungry and, often, poor artist, that historically fueled some of the philosophic underpinnings of some copyright systems  and, in particular, purportedly justified moral rights and duration extensions. 

Over the last century, the subject matter of copyright has seen an expansion as the proliferation of information has created "new"  subject matter capable of protection with the justifications for protection aligning more with traditional economic rationales of copyright and IP systems.  So what happens when a "modern" technology, Artificial Intelligence (AI), meets a field more closely associated with the romantic (big and little 'R') notions of copyright.  Last week's story from New York, sheds some light on these issues.  The AmeriKat has turned to Kat friend, Alexander de Leeuw at Dutch IP firm, Brinkhof, who reports on last week's news:

"Last Friday was a remarkable day in the history of art sales. For the first time a painting made entirely using ‘Artificial Intelligence’ was sold at auctioning house Christie’s New York. The painting with the title “Edmond de Belamy, from La Famille de Belamy” was originally estimated at $7,000-$10,000 but ultimately sold for the staggering amount of $432,500 (including fees). Much like a painting of human origin, the painting sold at Christie’s was signed by ‘the artist’ with a core component of the algorithm that created it:“min G max D Ex[log(D(x))]+Ez[log(1-D(G(z))]”.
The artist's signature? 
So how did this portrait of the imaginary Edmond de Belamy come into existence? It essentially comes down to teaching a computer how to paint. The portrait was made with machine learning algorithms called Generative Adversarial Networks (“GANs”). These algorithms are capable of generating images by mimicking characteristics of images from a training dataset (but are also capable of generating other output such as music and text). The artist Pierre Fautrel of the Paris-based art collective Obvious Art (which includes Hugo Caselles-Dupré and Gauthier Vernier) inserted 15,000 portraits dating from the 14th to the 20th century into the software, enabling it to make portraits itself. This led to a series of eleven unique images portraying the ‘Belamy family’, of which Edmond de Belamy is one.

The sale of Edmond de Belamy raises fundamental questions as to how centuries of IP law will interact with AI, the obvious question being whether there can be copyright in an AI-created portrait. Visually these paintings can hardly (if at all) be distinguished from paintings of human origin, supporting the view that if copyright protects expression (as opposed to an idea), there is an artistic expression to some extent. On the one hand, the AI can be seen as a tool used by a human artist. On the other hand, one could argue that the AI itself is emulating creativity, not the person feeding data to the machine. But is that enough, or even relevant for copyright law's purposes? Even assuming there can be copyright protection, a fierce debate can be held over who is the author and, therefore, who owns it. Is it Pierre Fautrel, who used the software to create the series of paintings and selected the paintings for the dataset, or are we in the realms of joint ownership with the creator of the algorithms? These questions will become more difficult to answer with the increasing potential of independently operating AI systems [particularly if AI datasets are pulling from open source material, notes the AmeriKat].
The answers to these questions can have far reaching implications, from everyday issues such as whether Edmond de Belamy can be freely copied by others, to more peculiar questions, for example, with respect to the droit de suite. With the fast-paced development of AI algorithms, their broad applicability, and the recent high-value auction of Edmond de Belamy, these questions will have to be answered. National courts are likely to grapple with these questions sooner rather than later.

Under Dutch law there is a rather low threshold for obtaining copyright protection. Hence, if presented in front of a Dutch court, copyright protection would likely be afforded as long as human choices – such as feeding specific information to the software – are involved. Even if no human choices are involved other than in creating the software, for example by connecting the software to Google images for its data input,  copyright protection would likely be afforded to the person who wrote the software, both for the software and the resulting painting.
A distinct IP question is what are people paying for when buying AI paintings? In legal terms they are buying a physical object. To the extent that any copyright exists (which may differ from territory to territory) they may also be getting an implicit or implied license. However, not all purchase agreements provide much clarity on what rights the buyer is getting. It is questionable whether amounts in the range of $432,500 would be paid for non-exclusive AI paintings. [Which creates a fun question as to whether, when Christie's delivers your painting, they also give you a hard drive with the code, says the AmeriKat].
A more policy driven question also needs to be answered:  Why are buyers willing to pay so much for AI art? Is it the concept of buying a new – and currently still rather unique – form of art?  In other words, is this a ‘first time value’? Or are people paying for the actual (artistic?) expression of the algorithms?"
Perhaps the answer to that last question may answer whether copyright subsists and should protect such works and which part of the subject-matter, muses the AmeriKat. If emotion has little to do with so many protectable works now, then why treat AI any differently?   Does it matter that this form of AI algorithm is using data sets to imitate previous work in creating a new work (i.e. GAN), as opposed to the CAN (Creative Adversarial Networks) system at Rutgers which is using data sets to create novel works - something different than the data?  Or are different AI systems just doing what artists have done for centuries - imitating and breaking the mold? 

On the final question of authorship, Caselles-Dupre explained that:
"If the artist is the one that creates the image, then that would be the machine. If the artist is the one that holds the vision and wants to share the message, then that would be us.  "
What do readers think?  
Auctioning Art(ificial Intelligence): The IP implications of Edmond de Belamy Auctioning Art(ificial Intelligence):  The IP implications of Edmond de Belamy Reviewed by Annsley Merelle Ward on Wednesday, October 31, 2018 Rating: 5


  1. There are a number of ideological arguments for not extending the standard copyright model to works entirely created by AI, not least of which is that the method of calculating the term cannot, realistically, be based on the lifetime of the 'author'. Perhaps a similar approach to the USA's corporate ownership term might be made to work. That said we already have the, to my mind ludicrous, situation where a piece of software can be protected for upwards of 120 years even though the hardware on which it was designed to run would probably become obsolte within 10 or 20 years. If legislators are equally intellectually moribund when the time comes to consider the issues involved in just extending standard copyright legislation to AI generated works, we will no doubt experience further such illogicalities.

    There is a another question to be addressed. Would it be necessary to distinguish between a work which although generated by AI, was only produced on the command of a human, and a work which was autonomously produced at will by the AI? Clearly the former state might be directly tied to the economic needs of the human, whereas in the latter case, the economic aspect is entirely absent from the motivation to produce the work. It would then follow that the economic model which underpins copyright today would not apply if an AI 'artist' operated autonomously.

    The other question - not raised in the guest piece - is whether the EU's Artist's Resale Right (EU Directive 2001/84/EC) could apply to AI- generated works. Clearly the Directive and its recitals were written with the implicit intention of rewarding human artists (or their heirs), but it doesn't explicitly exclude works generated by computers under human guidance, nor AI created works. With sums just short of half a million dollars at stake, this issue may arise rather sooner than we imagine.

  2. I simply don't believe Christie's when they say they are the first auction house to offer a work created by an algorithm. Maybe with some narrow definitions of both "algorithm" and "auction house"...

    Anyway I have always thought that this is the sort of thing the "computer-generated" work provision in s. 9 Copyright Designs and Patents Act was supposed to cover - and no I don't think 1988 was too early for it to be imagined. But it hardly covers all the bases. For example, who needs to provide the required originality? The person specified by s.9(3), or the computer? If originality is in line with Infopaq, Painer, etc. then it seems to me impossible for a deterministic machine to provide it. But if the originality comes from the person then s.9(3) could be meaningless, since in that case the author would probably be the creator of the work under s.9(1) anyway.

    It will be interesting to see if the paucity of case law on this provision continues for another 30 years. So far the lack of cases (unless someone can point me to dozens I've missed...) suggests that s.9(3) might indeed be pointless. Or perhaps now that some works of this nature are selling for $432,500, it will be worth someone's while to go to court over it.

  3. @Freddie Noble.

    With respect, I'm not sure that section 9(3) helps all that much. I think that the provenance of human controlled but computer-generated works is fairly widely accepted as protectable with copyright. There is a good deal of case law relating to the output from computer games, and of course the ubiqity of digital images which have been modified with software like Photoshop are certainly recognised as worthy of copyright protection. Indeed many other types of work including music, software, literature, architectural designs and so forth rely to a greater or lesser extent on computer-assistance in a way which probably wouldn't be possible (or economic) to create by human hand and brain alone.

    Section 9(3) was part of the original 1988 CDPA and as such represents something of a foresighted outlier amongst the copyright laws of the world. I think I am right in saying that no other EU member state provides such an explicit statment of inclusion, while the nearest the US Copyright Act gets to something similar is this within 17 USC Section 102: "original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device." Famously, works created by animals have been ruled ineligible for copyright in some US Districts.

    As the status quo seems to accept works which have been created either totally or in part by computers which are under the direct control of humans, I fear that without an adequate debate, it would be all too easy for that situation to evolve to cover works which effectively have no human input beyond hitting the on-switch. Because inforned public debate seldom has any bearing on court decisions, I suggest that the way forward with true AI genereated works should be through legislation not caselaw. However given the quality of the debate and the lobbying capacity of large multi-nationals seen with Articles 11 and 13 of the DSM, I suspect that even that route will get it wrong.

    Perhaps a post-Brexit UK can once again lead the way by enactng provisions to deal with this issue im a coherent way, before the courts have to deal with it for real.

  4. Part of the confusion originates from the strange way people treat results of a class of computer algorithms under the classification AI. These "AI" algorithms require inputs and the selection of the inputs determines the results. Before showing the results, there like has been a further selection process. The creative aspects are: conceiving the idea of using these tools, finding a set of meaningful inputs and filtering the results. Surely worthy of copyright protection.

  5. Andy, thanks for your comment. Let me expand a little.

    Section 9(3) ... something of a foresighted outlier - I agree with that. I think that's a large part of my original point, that UK copyright law (though only arguably, and not very clearly) tries to deal with this. Whether it succeeds is questionable.

    So I also agree somewhat with your comment that 9(3) "doesn't help" - given the problems it creates over the definition of originality which I raised in my first comment.

    You mention a good deal of case law on computer games. I know about Nova Productions v Mazooma, which says a little bit but not a lot. Is there anything else? Is there anything at all from the Court of Appeal or above or from anywhere which actually grapples with interpretation of this subsection?

    In general I disagree that most digital images modified with Photoshop are on point as computer-generated works. These works are human generated, in exactly the same way that a traditional painting is created by the artist and not by the paintbrush, a photograph is created by the photographer and not by the camera, etc. I think this because section 9 is all about authorship. s. 9(1) says that the author of a work is the person who creates it. In ordinary language you would say that a person is "creating a digital image using photoshop", i.e. using the computer as a tool to assist his creation. In relation to artistic works s.9 gives us nothing further, apart from in the case that the work is computer-generated where is says the author is "taken to be" the person who makes arrangements. So I read s.9(3) as applying to situations where it is just not accurate to say that any human "created the work" - potentially such as this AI example. In fact s.178 says this outright - computer-generated means there is no human author.

    So the problem as I see it is that a work to which 9(3) applies has no human creator, and yet must still be original. This problem arises even in the not-so-AI computer games cases like Nova. But maybe there is no contradiction, if what is required is for the computer to demonstrate originality. This is in line with some definitions of what AI actually is - "AI seeks to make computers do the sorts of things that minds can do" [1] (things like produce original art in the copyright sense?).

    Has that happened with this Belamy picture? I'm not convinced. As far as I can tell, the machine was deterministic. It exhibited no free will. It carried out a GAN algorithm with inputs provided by humans. The humans may not have known the result in advance but nevertheless the output was entirely controlled by human choices. There was no originality on the part of the machine. Is it really AI in the sense that the computer has done the sort of thing a [human] mind could do?

    Artificial intelligence has come some distance in recent years but, as far as I know, there is no a machine out there which can really demonstrate originality in the copyright sense.

    So if 9(3) doesn't apply for that reason, then can the three humans involved claim joint authorship under s. 9(1)? I would say probably yes. They may not have known the result in advance of execution but then, does a modern artist who splatters paint on a canvas?

    Whether the solution is by legislation or case law I am not sure. As you hint, it probably depends on our expectations as to the quality of the legislation and the quality of the case law. I for one do not know what to think the law should be in this area.

    By the way, Wikipedia have taken their usual view that the work must be in the public domain. Perhaps someone with a spare $432,500 will sue them.

    [1] M Boden; Artificial Intelligence: A very short introduction. There are many other definitions, some of them quite different.


All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.

It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.

Learn more here:

Powered by Blogger.