Copyright/Intellectual Property

European Parliament Approves Catastrophic Copyright Bill That Threatens the Internet

13 September 2018

From Gizmodo:

Members of the European Parliament voted Wednesday to approve a sweeping overhaul of the EU’s copyright laws that includes two controversial articles that threaten to hand more power to the richest tech companies and generally break the internet.

Overall, MEPs voted in favor of the EU Copyright Directive with a strong majority of 438 to 226. But the process isn’t over. There are still more parliamentary procedures to go through, and individual countries will eventually have to decide how they intend to implement the rules. That’s part of the reason that it’s so difficult to raise public awareness on this issue.

Momentum to oppose the legislation built up earlier this summer, culminating with Parliament deciding to open it up for amendments in July. Many people may have thought the worst was over. It wasn’t—but make no mistake, today’s vote in favor of the directive was extremely consequential.

The biggest issue with this legislation has been Articles 11 and 13. These two provisions have come to be known as the “link tax” and “upload filter” requirements, respectively.

In brief, the link tax is intended to take power back from giant platforms like Google and Facebook by requiring them to pay news outlets for the privilege of linking or quoting articles. But critics say this will mostly harm smaller websites that can’t afford to pay the tax, and the tech giants will easily pay up or just decide not link to news. The latter outcome has already happened when this was tried in Spain. On top of inhibiting the spread of news, the link tax could also make it all but impossible for Wikipedia and other non-profit educational sources to do their work because of their reliance on links, quotes, and citation.

The upload filter section of the legislation demands that all platforms aside from “small/micro enterprises” use a content ID system of some sort to prevent any copyrighted works from being uploaded. Sites will face all copyright liabilities in the event that something makes it past the filter. Because even the best filtering systems, like YouTube’s, are still horrible, critics say that the inevitable outcome is that over-filtering will be the default mode of operation. Remixing, meme-making, sharing of works in the public domain, and other fair use practices would likely all fall victim to platforms that would rather play it safe, just say no to flagged content, and avoid legal battles. Copyright trolls will likely be able to fraudulently claim ownership of intellectual property with little recourse for their victims.

. . . .

Joe McNamee, executive director of digital rights association EDRi, recently told The Verge, “The system is so complicated that last Friday the [European Parliament] legal affairs committee tweeted an incorrect assessment of what’s happening. If they don’t understand the rules, what hope the rest of us?” As we come closer to living parallel lives online and IRL, such sweeping legislation is dangerous to play with.

In a statement sent to Gizmodo, Member of European Parliament Julia Reda said, “Unfortunately, all the concerns by academics, experts and internet users that led to the text being rejected last July still stand.” She said that the EU is relying on “wishful thinking” rather than addressing the clear problems in the directive. Her overall assessment of the vote was blunt: “Today’s decision is a severe blow to the free and open internet. By endorsing new legal and technical limits on what we can post and share online, the European Parliament is putting corporate profits over freedom of speech and abandoning long-standing principles that made the internet what it is today.”

Link to the rest at Gizmodo and thanks to Kat, who points out that this could impact TPV (although it might be a “small/micro enterprise”), for the tip.

Although he claims no deep knowledge of the language of the legislation and its potential impact on TPV or elsewhere, PG says one possibility springs (or limps) to mind would be that it could create a sort of digital Dark Age for websites originating in the EU.

Who is going to quote or take the risk of linking to a European website that may trigger a “link tax” or otherwise expose the site owners to jurisdiction under a complex EU legal regime?

It is not difficult for PG to imagine a new WordPress plugin that automatically screens the IP addresses of linked articles or online visitors to a blog and blocks access to the linked website in general or for visitors with a European IP address. This type of digital censorship system will almost certainly be over-inclusive if the penalties involve exposing a US, Brazilian or Japanese blog to some sort of EU regulatory framework.

If the impact of such an initiative creates one or more digital “no-go” zones, the individuals and organizations in those zones could suffer a much greater harm than any offsetting benefit to the relatively small number of IP rights holders in any nation.

But, as usual, PG could be wrong. As mentioned, he only knows what he reads online about a wide variety of subjects these days.

While he believes himself to be a staunch advocate for the rights of authors and other creators to reasonable legal protections for their works, PG thinks a couple of items about the potential harms of rent-seeking are relevant.

From Wikipedia:

In public choice theory and in economics, rent-seeking involves seeking to increase one’s share of existing wealth without creating new wealth. Rent-seeking results in reduced economic efficiency through poor allocation of resources, reduced actual wealth-creation, lost government revenue, increased income inequality, and (potentially) national decline.

Attempts at capture of regulatory agencies to gain a coercive monopoly can result in advantages for the rent seeker in a market while imposing disadvantages on (incorrupt) competitors. This constitutes one of many possible forms of rent-seeking behavior.

. . . .

Rent-seeking is an attempt to obtain economic rent (i.e., the portion of income paid to a factor of production in excess of what is needed to keep it employed in its current use) by manipulating the social or political environment in which economic activities occur, rather than by creating new wealth. Rent-seeking implies extraction of uncompensated value from others without making any contribution to productivity. The classic example of rent-seeking, according to Robert Shiller, is that of a feudal lord who installs a chain across a river that flows through his land and then hires a collector to charge passing boats a fee (or rent of the section of the river for a few minutes) to lower the chain. There is nothing productive about the chain or the collector. The lord has made no improvements to the river and is not adding value in any way, directly or indirectly, except for himself. All he is doing is finding a way to make money from something that used to be free.

. . . .

Rent-seeking is distinguished in theory from profit-seeking, in which entities seek to extract value by engaging in mutually beneficial transactions. Profit-seeking in this sense is the creation of wealth, while rent-seeking is “profiteering” by using social institutions, such as the power of the state, to redistribute wealth among different groups without creating new wealth.

And some relevant quotes:

I readily acknowledge that it may be difficult to know where to draw the line between ‘corruption’ and ‘rent-seeking behaviour’…. The latter term is generally used to refer to the process by which interest groups adopt (lawful) means to secure competitive advantages from the political process and is a phenomenon widely recognised as influencing law and legal institutions in industrialised societies and is the subject of a huge literature . . . . Rent-seeking may, indeed, impose costs to the economy as high, if not higher, than those arising from corruption (narrowly defined).

~ Anthony Ogus

Time, talent, money, knowledge, and resources that could be used to protect people and cure deadly diseases are instead being wasted, siphoned off by rent-seeking because the government needs to…protect us.

~ Lisa Casanova

As soon as the state takes upon itself the task of planning the whole economic life, the problem of the due station of the different individuals and groups must indeed inevitably become the central political problem. As the coercive power of the state will alone decide who is to have what, the only power worth having will be a share in the exercise of this directing power. There will be no economic or social questions that would not be political questions in the sense that their solution will depend exclusively on who wields the coercive power, on whose are the views that will prevail on all occasions.

~ Friedrich August Hayek

Copyright and tattoos: where are we now?

8 September 2018

From The IPKat:

Copyright and tattoos is one of The IPKat’s favourite topics. Over time a few posts have been devoted to exploring the intricacies of copyright law as applied to this type of works.

. . . .

[T]attoos are artistic works with no particular features, if not the medium they are attached to, ie the human body. From a copyright standpoint, issues might arise in a number of circumstances, including when:
  • The owner of the body (clearly a tangible medium) on which the tattoo is affixed is not also the owner of the copyright in the work (the tattoo);
  • The tattoo reproduced on someone’s body is an infringing copy of a third-party copyright work.

. . . .

Scenario A – Owner of the body on which the tattoo is affixed is not also the owner of the copyright in the work

This scenario might not appear particularly problematic to the lay person, but it has been causing celebrities and athletes more than an headache over the past few years. Readers will remember for instance the (in)famous case of the Mike Tyson’s face tattoo and the lawsuit that his tattoo artist brought in the US against the producers of The Hangover – Part II over the unauthorized reproduction of that artistic work. The case was eventually settled out of court, but it was sufficient to generate a renewed new awareness around the legal issues potentially affecting tattoos.

This has become true also for sports organizations. For instance, in 2013 Forbes reported that the NFL Players Association (this is the organization that represents professional American football players in the National Football League) had become increasingly concerned about potential copyright claims concerning its members’ tattoos, and started advising agents to tell their players that, when they get tattoos, they should get an assignment from the tattoo artist and, if they can track down their former artists, they should get an assignment as well.

One of the reasons as to why such concerns have started arising is also because athletes routinely license use of their likeness in advertising, merchandising, and important ‘side’ products like videogames. And videogames seem to have become problematic indeed. 
As The IPKat reported back in 2013, videogame publisher THQ was sued by a tattoo artist over the UFC Undisputed games for reproducing, without his permission, the drawing of a lion that he had tattooed on the ribcage of mixed martial fighter Carlos Condit
 
Now, a new lawsuit – still concerning videogames and tattoos – is pending before the US District Court for the Southern District of New York against Take Two over its NBA 2K videogame franchise. The owner of copyright in a number of tattoos featured on the bodies of NBA basketball players, including LeBron James, is arguing that it is an infringement of its copyright to reproduce these artworks in the videogame without permission. Also James has weighed in on this issue, summarizing better than any lawyer what the real conflict – besides all legal technicalities – is in a case like this:

 In the fifteen years since I’ve been playing professional basketball, this case is the first time that anyone has suggested to me that I can’t license my likeness without getting the permission of the tattooists who inked my tattoos. No tattooist has ever told me I needed their permission to be shown with my tattoos, even when it was clear I was a public basketball player.

Link to the rest at The IPKat

Photographers, Don’t Think You Need to Register Your Copyrights? Think Again…

6 September 2018

From Medium:

So many people get caught up in the fact that “you don’t have to register the copyrights to your work because it’s copyrighted the moment you create it.” True. Anything you create is protected by copyright the moment you create it. HOWEVER, if you want real protection and legal hold when it comes to your work being stolen (copyright infringement), it’s worth the few minutes and few dollars to register it.

Last spring when I discovered eight of my own images had been copyright infringed, I became particularly passionate about the legal rights of artists. I was fortunate that I had done a few things to help with protecting my work. One step was including a copyright statement, which is clearly stated, on my website. But, the most important step I took was not only registering my copyrights, but doing so in a timely manner. As such, several of my images that were stolen had already been registered with the U.S. Copyright Office upon infringement. I can attest that it was worth the few pennies per image to do so.

For example, the fee for me to register 184 photos for 2013 and 134 photos for 2014 was $55 per batch. That was done via an online registration form with the U.S. Copyright Office. Note that the limit per batch is 250 photos for $55 and batched by calendar year. The $110 to register several hundred photos was the best insurance policy I’ve ever bought. I will also add that the people I spoke to and e-mailed with from the U.S. Copyright Office were incredibly helpful and knowledgeable.

. . . .

HH: Why should an artist copyright register their work?

CW: When a photo is not registered with the U.S. Copyright Office prior to the infringement, or within three months of the first publication of the photo, a copyright owner may recover only “actual damages” for the infringement (pursuant to 17 U.S.C. 504 (b)), instead of statutory damages. Courts usually calculate actual damages based on your normal license fees and/or industry standard licensing fees. You also may recover the profits the infringer made from the infringement if they aren’t too speculative.

If your photo is timely registered for an infringement, you will be eligible for statutory damages of up to $150,000 for a willful infringing use. See 17 USC §504(b) and ©. You also may be able to recover some of your legal fees and court costs from the infringer. See 17 USC §505. Additionally, you need to have received your registration certificate to file a complaint for a copyright infringement lawsuit in most federal courts.

HH: With so many artists posting their work online, how important is it to use the © symbol, and what is the proper format?

CW: Including a copyright notice is no longer required for copyright protection, but it is a good idea to use it. When the copyright notice is used, it may stop someone from stealing your art, either because it serves as a reminder that the work is protected or because the notice interferes with the use of the work. When you post a copyright notice along with your registered images, then the infringer cannot claim that the infringement was innocent (reducing the damages to as low as $200 per work) and a court is more likely to find that the infringement was willful, supporting the maximum in infringement damages.

The official copyright notice has three parts: the first part is the © (the letter “c” in a circle), the word “Copyright,” or its abbreviation, “Copr.” The second part notes the year when the work was first published. The third required part of a copyright notice is the name of the copyright owner. For example, the final format looks like this: © 2014 Carolyn E. Wright.

Link to the rest at Medium

PG says other than the option to submit multiple photos for registration, the rest of the OP will apply to authors as well as photographers.

Universal Music Settling Big Class Action Lawsuit Over Digital Royalties

6 September 2018

From Billboard:

An important chapter in the legal history of the music business may be coming to conclusion soon as Universal Music Group is close to submitting a settlement resolving claims that it cheated recording artists of royalties from digital downloads.

The putative class action from artists including Chuck D. of Public Enemy, Rick James (by way of trust), Dave Mason of Traffic, Whitesnake, Andres Titus of Black Sheep, Ron Tyson of The Temptations, among others, alleges that record labels should be treating digital download income off of venues like Apple’s iTunes as “licenses” rather than “sales.” By accounting the other way, the artists get about 15 percent of collected income rather than 50 percent they allege is due.

. . . .

The monetary value of UMG’s coming settlement haven’t yet been disclosed, but The Hollywood Reporter has learned that it will likely cover EMI, which was acquired by UMG in 2012 and has been dealing with its own litigation on the digital download front. Following settlements by Warner Music and Sony, UMG’s deal if approved would mean that all of the record majors have resolved claims following the 2010 appellate ruling in F.B.T. Productions v. Aftermath — dealing with Eminem songs — which suggested that “licenses” rather than “sales” were the more appropriate accounting treatment in an era where record labels no longer spend huge amounts on packaging physical CDs.

According to sales data released this week by the RIAA, digital downloads is the top revenue producer in the music industry. Download sales are at $2.64 billion, which beats physical music sales of $2.27 billion.

. . . .

As streaming gets closer to dominating downloads, the litigation may shift likewise as well.

The digital downloads cases may be on the precipices of conclusion, but scrutiny may follow as to whether record labels are cheating artists on money collected from outlets like Spotify. For instance, on Tuesday, a judge refused to reject claimsthat Sony breached agreements and good faith dealing with 19 Recordings — the label of former American Idol contestants Kelly Clarkson, Carrie Underwood and Jordin Sparks — by allegedly mischaracterizing income from streaming services as as “sales” or “distributions” rather than as “broadcasts” or “transmissions.”

Link to the rest at Billboard

Starting several years ago, PG blogged about this issue before in the context of book publishing agreements. For a great many years, standard publishing contracts included royalty provisions that paid a much lower royalty for books that were sold than for books that were licensed.

PG hasn’t surveyed Terms & Conditions for online ebook sellers for a long time, but the last time he did, the major online retailers all said something to the effect that the publishers were licensing ebooks to readers, not selling them.

Here’s the relevant portion of Amazon’s current Kindle Store Terms of Use, located in Paragraph 1, “Kindle Content”:

Kindle Content is licensed, not sold, to you by the Content Provider.

If you pull up a publishing contract that’s more than ten years old and search for the royalty rates payable when the publisher licenses the book, you may well find that, like the music business, Big Publishing paid much higher royalties for licenses than it did for sales.

Appeals Court Revives Lawsuit Against CBS Over Pre-1972 Recordings

5 September 2018

From The Hollywood Reporter:

CBS Radio must again contend with a lawsuit brought by ABS Entertainment, owner of recordings by Al Green and others. On Monday, the 9th Circuit Court of Appeals wiped out an eye-opening 2016 summary judgment ruling by a trial court in CBS’ favor that raised the prospect that music owners could enjoy perpetual copyright because remastered versions were independently copyrightable. In the decision, the appeals court concludes that the judge shouldn’t have ruled so quickly for CBS and casts doubt on whether remastered sound recordings exhibit enough originality to be copyrightable.

For decades, radio broadcasters didn’t have to compensate sound recordings owners for performance.

But then, owners of older recordings — those authored before Congress conferred federal copyright protection for sound recordings — asserted claims of misappropriation under state laws. Like SiriusXM and Pandora, CBS Radio has faced a wrath of litigation for broadcasting iconic songs from the dawn of the rock era.

In response to the lawsuit, CBS made a bold and novel argument. The broadcaster made the point that older recordings were mainly distributed through vinyl, and what the broadcaster now performs is really remastered versions, eligible for its own copyright protection. As such, any state misappropriation claim is preempted by federal law, and owners of pre-1972 recordings don’t have any claim for compensation.

. . . .

“We conclude that the district court erred in finding a lack of a genuine issue of material fact about the copyright eligibility of remastered sound recordings distributed by CBS and improperly concluded that ABS’s state copyright interest in pre-1972 sound recordings embodied in the remastered sound recordings was preempted,” he writes.

Judge Linn acknowledges that derivative works do enjoy independent copyrightability, but focuses the question on originality and whether the author of a derivative work has contributed something that is creative and easily recognizable as distinct from the original. He points to guidance from the U.S. Copyright Office that mere changes in format, declicking and noise reduction don’t warrant separate copyright protection.

. . . .

“If an allegedly derivative sound recording does not add or remove any sounds from the underlying sound recording, does not change the sequence of the sounds, and does not remix or otherwise alter the sounds in sequence or character, the recording is likely to be nothing more than a copy of the underlying sound recording and is presumptively devoid of the original sound recording authorship required for copyright protection,” states the opinion . . . . “Such a work lacks originality.”

. . . .

The trial judge is faulted for placing too much reliance on CBS’ musicologist expert who attempted to explain how listeners of remastered versions would perceive changes to timbre, spatial imagery, sound balance and loudness. The 9th Circuit derides these bits as “technical improvements,” and while devoting some space to credit the contributions of recording engineers and producers, the panel of judges are leery of extending copyright too far.

Link to the rest at The Hollywood Reporter

In addition to the obvious copyright law implications, the OP reminded PG of the sometimes substantial difference between the copyright attitudes of West Coast creative individuals and the organizations they work with and those of comparable (at least with respect to copyright) East Coast individuals and organizations.

West Coast creatives (and their lawyers) tend to be much more persnickety and obsessive about the nature of copyrights and the scope of rights granted by a copyright holder. This is why West Coast copyright disputes – often related to music and movies – tend to be litigated with some frequency (by intellectual property standards).

OTOH, East Coast creators – often authors of books or other writings – and their advisors tend to be more blasé about grants of rights based upon their copyrights.

For example, the typical Big Publishing publishing contract grants extremely broad rights to the publisher for the full term of the copyright with an extraordinarily low bar for the publisher’s reciprocal obligations to the creator. If the publisher pays the agreed-upon advance and publishes a copy of the author’s book, the publisher has essentially locked up a wide range of rights to the book and derivative works arising therefrom for the rest of the author’s life plus 70 years in the US.

PG has used geography to distinguish between two different habits of copyright-based businesses, not to disparage anyone for where they live or work, but rather to contrast what have become the common practices with respect to creators in the movie and music industry (West Coast) and those in the traditional publishing industry – books and periodicals (East Coast).

While differences in West Coast media and East Coast media are substantial and the comparison is not completely fair, PG notes that typical working creatives involved in West Coast copyright-based industries are likely to earn more money from their creations than typical working creatives in East Coast industries relying on copyrighted creations.

While starving screenwriters are not hard to find in Los Angeles, starving authors are also common in New York.

But PG could be completely wrong. Commenters are, as usual, free to dissent or identify factors PG has foolishly overlooked in his maunderings.

Everybody Dance Now! Actually, Don’t … That Choreography May Be Copyrighted

5 September 2018

From Trademark and Copyright Law:

While an amazing piece of dance can make everyone want to join in, there are some dances no one should mimic without authorization, at least not as part of a commercial “public performance,” as that term is defined by 17 U.S.C. § 101. Why not? It’s not because they won’t look good performing the dance (we remain agnostic on that), but because the choreography for the dance may be under copyright.

. . . .

Although dance is an art form as old as time, copyright protection for choreography in the United States is a relatively recent development, created when the 1976 Copyright Act added choreography and pantomime to the list of copyrightable subject matter. Choreography, as opposed to dance itself, is the composition and arrangement of “a related series of dance movements and patterns organized in a coherent whole.”

Prior to 1976, some of the dances from modern or contemporary companies might have been left out in the copyright cold. Before the 1976 Copyright Act added choreography as its own category, works of dance could only be copyrighted if they fell within the definition of a “dramatic work.” This category had some requirements that might not be met by much modern dance (for example, the requirement that the work develop characters or express an emotion through specific movements). Since the 1976 Act, modern and contemporary dance is more likely to be afforded copyright protection because no clear narrative arc is required so long as the choreography “convey[s] some thematic or emotional concept to an audience.” The timing of this protection puts the U.S. behind other countries, like France, where the first copyright protection for choreography was established in 1862.

. . . .

Because all copyrighted works must be “fixed in a tangible medium of expression,” choreographic works are typically submitted to the Copyright Office for registration on video; or described in text or depicted in photographs or drawings; or recorded in a traditional dance notation format (e.g., Labanotation or Benesh Notation). Regardless of which of these forms is used, the fixation must be in sufficient detail to be the template for subsequent, uniform performances.

When considering whether a work is “choreography,” the Copyright Office considers several factors typical of choreographic works, including the presence of rhythmic movements from a dancer’s body in a defined space; compositional arrangement into a coherent, integrated whole; musical or textual accompaniment; and dramatic content such as a story or theme. The Copyright Office will also consider whether the work is designed to be performed by skilled individuals for an audience. Works must additionally have a sufficient level of original authorial contribution, which means that they must be independently created and contain a sufficient amount of creativity.

Link to the rest at Trademark and Copyright Law

‘Perma.cc’: is the fight against “link rot” copyright compliant?

1 September 2018

From The IP Kat:

From news outlets to academic writing, publishing online is now part of the mainstream amongst publishers. It is relatively inexpensive, instantaneous and reaches readers worldwide. But the dynamism of internet publications does have one inconvenient– “link rot”.

‘Link rot’ refers to the decoupling of the hyperlink (or URL) with the webpage with which it was originally associated, rendering the link useless. While you may not be familiar with the phrase link rot itself, undoubtedly you will have experienced some of its most irritating symptoms: ‘page error 404’, ‘The URL you requested was not found’ or ‘Oops! Something wrong happened’. Research shows that, on average, a staggering 50% of links will be decoupled from their original content, i.e. turned to rot, two years following publication.

. . . .

With weblinks spreading to footnotes, research paper citations and court reports referencing evidence or information, link rot has become a threat to academic rigour (in addition to being a source of acute frustration). In short, link rot has become an evil in need of a cure.

Technically, the remedy for link rot is rather simple: permanent links (also known as ‘permalinks’). The publishers of the content can attribute a permanent (inseverable) link to the webpage. According to TechnoPedia,

permanent links work by providing an alternate but permanent Web address for content, which is initially viewable only on the home page or top-level domain … but is relocated to a separate page once it’s archived.”

But it is for the publisher of the content to take this precaution, and not all dynamic website hosts or editors care, or have the wherewithal, to preserve their publications from link rot.

Harvard’s Library Innovation Lab addressed this problem by creating a tool that enables the content user, in lieu of the publisher, to take the necessary precautions to preserve the webpages they reference. This tool goes by the name of ‘Perma.cc’ and it is a free service presented as the antidote for link rot in research and legal scholarship. Interestingly, this tool was first designed to address the issue of link rot in law journals, which is claimed to affect 20% of online material after two years, reaching 50% after five years.

. . . .

Perma.cc enables a user of online content to directly create a permanent link (as opposed to relying on the website editor) through its platform. This permalink is associated with an identical copy of the webpage that the user would like to preserve. The copy of the webpage is then moved to another page, an archive page, hosted by Perma.cc. The ‘archive’ copy of the page is given a new, permanent link, which can then be cited in a paper, brief or court decision, without fear of rot.

. . . .

The benefit of user-generated permalinks, which Perma.cc facilitates, is undeniable. But is it compliant with copyright law?

Webpages invariably feature copyright-protected content. Platforms which make the most of what the online medium has to offer will not only have text-based content but also embed images, videos and sound. So what happens when all of this content is captured by Perma.cc to be rehosted on an archive page and given a new weblink? Is this of infringement under copyright? Is it a form of reproduction or communication to the public as we understand each of these provisions under the copyright law? Would perma.cc benefit from an exception to copyright protection?

At present, it seems that Perma.cc has been endorsed and used by mainly US-based institutions. US copyright law include a fair use doctrine that is widely considered to provide generous protection for users of protected works in the furtherance of the public interest (as Perma.cc does with respect to research and information by maintaining accurate records of online sources).

By contrast, the scope of the fair dealing doctrine in the UK, or in most other European jurisdictions, is more circumscribed.

Link to the rest at The IP Kat

Here is the Permalink for this post – https://perma.cc/D5SX-ARNG

Can a Work of Art Created by AI be Protected by Copyright?

1 September 2018

From The 1709 Blog:

We now learn that auction house Christies will sell in October a work of art titled     [  ())] +  [( − (()))], Portrait of Edmond de Belamy, from La Famille de Belamy, which was created using AI.

Portrait of Edmond de Belamy – used for discussion purposes only

It is the work of a Paris-based collective, obvious art, founded by Pierre Fautrel, Gauthier Vernier and Hugo Caselles-Dupré. The work was created using Artificial Intelligence (AI) technology, more precisely the Generative Adversarial Networks technology invented in 2014 by Ian Goodfellow, which can create images. The name of the work, Edmond de Belamy, is an homage to Ian Goodfellow, whose last name can be translated in French as “Bel ami.”

Obvious art created a program, fed it with information about some 10,000 portraits from the 15th to the 19th Century, and Edmond de Belamy was printed. The whole process is explained on obvious art’s website.

. . . .

The portrait shows a man painted over a black and gray background from which he appears to emerge, dressed in black, with a white collar, in a fashion reminiscent of 17th century Dutch paintings. His features are not precisely lineated and one does not even see his nose. He is looking at us from an angle, and appears to have been painted by large brushstrokes.

The collective’s goal was to prove that machines can also be creative, just like humans (see thisinterview in French). It is an algorithm which created the work. Does that mean that Edmond de Belamy cannot be protected by copyright?

. . . .

Obvious art used the formula of the loss function of the original GAN model as the signature for the painting they created. If a program is the author of the work, then the work cannot be protected by copyright.
The U.S. Copyright Office clearly stated in its Compendium of U.S. Copyright Office Practices that a work must be created by a human being to be protected by copyright, and that“[t]he Office will not register works produced by nature, animals, or plants, giving as an example a work which cannot be protected by copyright a “photograph taken by a monkey.” It could now add “a painting created by AI.
. . . .
But even if we consider AI to be a mere tool, it is not a tool like a spade or a brush, as this tool had to be created and could be protected by copyright.
Computer programs can be protected by copyright and software, as it is a computer program, can be protected. A software is defined by the copyright Act as “a set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result” and U.S. courts use an “abstraction-filtration-comparison test” [see here for example] to find out which elements of a computer program can be protected by copyright.
If the tool used to create a work is protected by copyright, does that mean that the work thus created is also protected? Not necessarily, as Section 721.6 of the Compendium specifies that “ownership of the copyright in a work is distinct from ownership of any material object that may be used to create that work. The fact that the author used a computer to write an article, short story, or other nondramatic literary work does not mean that the work is a computer program.

Link to the rest at The 1709 Blog

« Previous PageNext Page »