From March to September 2016 the team is joined by Guest Kats Emma Perot and Mike Mireles.

From April to September 2016 the team is also joined by InternKats Eleanor Wilson and Nick Smallwood.

Sunday, 31 May 2015

3D printing and the law: three recent studies -- and a critical review

Last week, in "3D printing and the law: three recent studies and some recommendations", here, this weblog posted news from Dinusha Mendis of three recently-published reports on the intellectual property implications of 3D printing. The reports (all of which can be accessed here) arose from a project commissioned by the UK Intellectual Property Office (UKIPO).  To refresh readers' memories, they are (i) Study 1: A Legal and Empirical Study of Online Platforms and an Analysis of User Behaviour, (ii) Study II, The Current Status and Impact of 3D Printing Within the Industrial Sector: An Analysis of Six Case Studies and (iii) an Executive Summary.

This Kat hasn't yet found the time to read them -- but fellow 1709 Blogger Andy Johnstone has. Wearing his critical glasses, Andy writes as follows:
I found Section A of Study I -- the legal analysis part of the study -- rather disappointing. If 3D printing is the disruptive technology some claim it to be, in what way does it or will it disrupt IP law in the way that Lipson and Kurman predict in their forecast that "intellectual property law will be brought to its knees", as quoted in the Introduction to Study I?

What LucasFilm was all about
Section A really doesn't address the wider issue. Surely the first question is, which areas of IP law should we be examining in relation to the end product: is it copyright, design right, trade mark or patent law which is most affected? The authors have effectively discounted all but copyright, even though they then show that, following such cases as the UK Supreme Court decision in Lucasfilm [noted by the IPKat here], copyright will not normally apply to practical articles. Because of the over-reliance on copyright there is not a single mention in Section A of the equally, if not more, relevant UK and EU law on designs (ie the 'Designs' bit of the Copyright, Designs and Patents Act 1988 (CDPA), the Registered Designs Act 1949, the EU Designs Directive 98/71 and the Community Design Regulation 6/2002).

This perhaps occurs, in part, because Section A of Study I fails to set itself a clear aim. At one point (paragraph 1 page 2) the authors speak of "further interesting questions about the intellectual property status of such design files ..." and later (third paragraph Page 3) they say; "this Report aims to provide a clearer understanding of 'how' the sharing happens ... " and then later (paragraph 4 Page 30): "Section A of this Report will begin with a consideration of the copyright implications arising from the access and use of these online platforms." The last sentence of paragraph 3 Page 3 states that the Report will focus particularly on the copyright element for "reasons as explained below" but then fails to provide any reasons.

The Section's second weakness, to my mind, is that it too readily conflates the design drawing with the final 3D object. Much time is spent agonising over whether a CAD file can be subject to copyright if the final product doesn't fall neatly into a copyright category. A better approach would be to examine the two aspects in isolation, much as one might see the copyright in a song as being entirely separate to the copyright in the sound recording of the song (the phonogram). If design right law had been used as the vehicle for examining the issues, life might have been a little more straightforward.

CAD Cat here
But after a less than focused Introduction, the Report becomes almost immediately sidetracked into whether a Computer Aided Design (CAD) file is protected by copyright, without analysing which intellectual property rights are at risk through the introduction and development of 3D printing. Why, for instance, is the copyright status of the design file seen as being so important? Two completely different CAD files could produce exactly the same finished article, just as two separate sets of architects' drawings produced independently of each other could result in two houses which looked substantially the same. If the CAD file has been copied without permission, is the legal issue about infringement of the file, or about the production of the same end product?

The Report, having failed to answer that question about whether or a CAD file is protected by copyright, then addresses licensing, and whether exponents of 3D printing know about their rights and how to protect them. Neither problem arises solely from the advent of 3D printing; indeed almost exactly the same comment could be made about users of Facebook and their photographs, so in this respect Section A tells us nothing about the specific legal implications of 3D printing. Happily, on a positive note, I can say I entirely agree with the Report's recommendation that nothing needs to be done right now on the legal front -- but for entirely different reasons.

A more detailed look at the issues

Surely it's about more than just copyright

Presumably the main evidence of any sort of infringement, if it exists, is to be found in the end product -- the thing which the 3D printer makes. But by starting with the design document (ie the CAD file) the authors go up a blind alley because, apart from the technical process by which the item is manufactured in 3D printing, there is no outward difference between a CAD file and a pdf or a mpeg, or a jpg or an avi file; they are all digital files which define, respectively, a 3D object, a text document, some sounds, an image, or film/video once they have been converted into a form perceptible by a human being. There is little question that the current law recognises a jpg file as both a means of recording a work and as a copy as far as it concerns copyright in the artistic work it defines, just as an avi file is taken to be a copy of the video it holds in digital format. The CDPA already says (s.17(2)) that 
"Copying in relation to a literary, dramatic, musical or artistic work means reproducing the work in any material form. This includes storing the work in any medium by electronic means." 
Thus if the final product is one which falls into the categories of protected works, the intermediate digital form, be it CAD or jpg, is clearly protected. But what if the product is too utilitarian or practical to qualify as an artistic work? We know from s. 9(3) that 
"[i]n 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" 
and so the law seems relatively clear that a work generated by a computer under human guidance (which is what a CAD file is) can certainly be protected by copyright, but is the work which the CAD file represents an artistic work? Given that it bears many similarities with an architect's drawings (incidentally these are almost exclusively drawn by CAD these days), which are explicitly protected, I think the answer has to be that the design for anything will be an artistic work provided it meets the basic originality test applicable to any work; to that extent it is immaterial whether the end product is utilitarian or artistic, just as the building constructed to the architect's design can be artistic or utilitarian, without affecting the copyright status of the architectural drawings.

Who says a CAD File is a computer program

The major mistake I believe that Section A makes is trying to make a CAD file fit into the definition of computer program. This tends to suggest a complete misunderstanding of 'computer program' as defined in the software directive, and more generally how the term is understood by information technology specialists. Evidence of this misunderstanding can be seen in the assertion in paragraph 3 on page 9 that the CDPA 
"... states that a computer program and its embedded data for example is recognised as a literary work under copyright law" 
with a footnote reference to ss 3(1)(b) and (c) CDPA. From this the authors go on to state that 
"... it can be argued that a computer program encompasses an object design file or CAD file within its definition and is therefore capable of copyright protection as a literary work". 
The problem with that argument is that s. 3(1) CDPA makes no mention of embedded data. What it actually says is that a literary work includes "(b) a computer program; [and] (c) preparatory design material for a computer program." This form of words is taken directly from Article 1 of the EU Software Directive 2009/24. There is no mention of embedded data, or indeed any other form of data, which is what CAD file comprises. Just to be clear, the reference to 'preparatory design material for a computer program' is referring to flow diagrams, systems analysis and other precursors to writing a program, not design drawings. No computer specialist would ever confuse the terms 'computer program' and 'data'. Regrettably a lay person might. 
One consequence of this line of argument is that, by only considering CAD files in terms of literary copyright (because that is the category under which programs come) the Report fails to consider a CAD file in terms of the type of work it represents, namely a 'design', which will almost always means it will be an artistic work, irrespective of whether the final object attracts copyright. 
To be fair to the authors of the Report, they are often merely quoting others who seem equally confused about the nature of CAD files and want to treat them as 'software'. And at the foot of page 10 the authors almost concede that a CAD could be an artistic work. But it's a single sentence which is not pursued, before reverting on page 12 to the former viewpoint of seeing CAD files solely as computer programs. But let's just be perfectly clear about this, a CAD file is just a structured string of data which needs to be interpreted by software in order to render it either onto a screen or into a 3D physical object. It is no more complicated than Photoshop .psd file which can contain both the basic image taken by a digital camera and a record of all the alterations made to the image during the editing process. Or to take another example of a '3D' file, let's look at the myriad of video file formats in existence. These all store data covering three basic dimensions - x and y axes data, and time - plus a sound track, and occasionally such things as sub-titles and alternative camera angles. There is little doubt that if someone illegally copies a digital video recording such as that found on a DVD or BluRay, that amounts to infringement of the original work, and counsel for the plaintiff would not need to spend a moment in convincing the court that this digital recording was undoubtedly a 'work' even though no-one in the court could see or hear it at the time the physical DVD itself is exhibited. Not until it is placed in the correct player will the video be perceptible to human senses. Exactly the same principles apply to a CAD file. Load the CAD file into a suitable computer running the requisite software, and the design(s) held within the CAD data will be made visible.
Modify someone else's or scan your own file

The Report briefly touches on the law which might apply to modifying a design -- something with which I suggest the current jurisprudence is perfectly well equipped to deal. There is a suggestion that some of the apps which the 3D 'platform' websites make available might represent game-changers in terms of the modifications users are thus enabled to make to existing designs. Fortunately this is not taken further, but suffice it to say that 3D printing technology and these apps offer absolutely nothing new that we have not experienced in the past 25 or so years that Microsoft Word (for text), Photoshop (for images) and Qubase (for music) have been available. If the specific issue is with cheap apps, then these too have been available for several years via sites such as Instagram, and more recently Apple's AppStore, and they have been used widely for image manipulation during that time. The Report then looks at the copyright implications of scanning an existing object in order to create a new design file. Once again we have plenty of settled law about creating derivative works, and what might constitute additional intellectual creativity reflecting the personality of the scanner, during the post scan editing process. The examples given of current practices in the 3D printing world do not raise any new threats to the existing law.

Design Right

So what would adding an analysis of current design right law have added to the Report's scope? Quite a lot really, since design right is specifically intended to protect the sorts of objects most likely to be produced by 3D printing. The principal advantage is that by referring to design right law, the focus moves away from the (in my view, artificial) problem about whether or not a CAD file is subject to copyright, and moves it on to the protection available for the end product. Under design right law what matters is not whether the actual design document has been copied (that can continue to be dealt with under copyright law) but whether the product has an individual character (cf 'originality' under copyright law) and whether the 'copy' produces the same overall impression on an informed user. This answers the 'problem' raised in the fifth paragraph on page 14 where Interlego v Tyco Industries is cited concerning practical or utilitarian items not covered by copyright. It also clarifies the status of scanned products mentioned in the first paragraph of page 15. The second advantage of design right is that it makes no reference to the material from which the item is made. Chocolate, human tissue and organs -- all mentioned as 'issues' raised by 3D printing -- remain just materials which are irrelevant to the test for copying of a design. Arguably, the reproduction of human tissue and organs is far more likely to run into ethical problems than it is to lead to infringement claims.

Do we have a regulation problem?

Finally the Report examines 'regulation' of the various platforms which represent the major user communities where 3D printing is being developed and exploited. Apart from the licensing issue already referred to, the main concern here seems to lie in the fact that most of these platforms operate from within the USA and so, the Report contends, this poses jurisdictional problems for the way that infringement issues can be resolved. The Report examines what the terms of service of the various sites have to say on the matter, but once again we find that no new concept within IP law is engaged by 3D printing, which the courts have not already had to address with 'old' technologies such as BitTorrent, streaming video, or satellite TV for that matter.

I can't help thinking that when Professor Hargreaves advocated that changes to IP policy should be based on hard empirical evidence, he possibly didn't have this sort of tour d'horizon in mind.


Anonymous said...

The author parses too finely.

Software includes CAD.

"Data" defined is where the devil of the details awaits.

Not all data is equal (in the Orwellian sense). The U.S. captures this notion in its ability to ascribe patentable weight to any writing that is "functionally related." Thus, whether or not something can be classified as "data" is a misdirection, as it is not important at all whether the item is "data," but rather, it is how that item interacts with its environment. It is more than just a "load the right player" situation (for one thing, the parametrics involved in today's CAD packages are more than simply "pretty pictures."

I think that the overall picture is better put in that any one "object" may very well have multiple aspects, of which these multiple aspects may garner different protections under multiple laws. These are not mutually exclusive and any attempt to treat "objects" as being mutually exclusive simply fails at the start.

Andy said...

@anon 14.20.
I was at pains to distinguish between CAD as a process (which is achieved using software, or as the UK legislation terms it, 'computer program') and a CAD output file which is most definitely data. The authors of the study used the term CAD object design file and so I have no hesitation in maintaining my position that trying to examine a CAD file as if it was software (and thus a work of literature) is the wrong approach.
With all due respect to American jurisprudence, since the study was conducted at the behest of the UK IPO, the context in which the data / software issue has to be decided, initially at least, is the EU one. Because the authors of the study chose to omit patents as a category of IP which might apply to 3D printing (that is, the output process, as opposed to the means), we can thankfully leave on one side the familiar issue that the US and EU have chosen different paths when it comes to the extent to which software is patentable.

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':