Are Engineered Genetic Sequences Copyrightable?: The U.S. Copyright Office Addresses a Matter of First Impression

The Office's assertion that engineered genetic sequences do not fall within an explicitly enumerated category

The Copyright Office's primary argument against registering engineered genetic sequences appears premised on its assertion that genetic sequences do not fall within any of the eight categories of copyrightable subject matter explicitly enumerated in Section 102(a) of the Copyright Statute, and that neither the Office nor the courts have the authority to declare copyrightable subject matter that does not fall within one of these eight categories. But this interpretation of the law is entirely inconsistent with both the language and the legislative history of the statute. While it is true that Section 102(a) enumerates eight categories of copyrightable subject matter, including, for example, literary works, musical works, and dramatic works, it explicitly states that “works of [copyrightable] authorship include the [enumerated] categories.” In other words, the eight enumerated categories are illustrative, not limiting, and a work does not necessarily have to fall into one of these categories to be afforded copyright protection. This has long been the conventional understanding of copyright attorneys and scholars. For example, the leading copyright treatise, Nimmer on Copyright, explains that it is “clear that ‘works of authorship’ are not necessarily limited to the eight broad categories of works listed under Section 102(a).” ¹⁴ Nimmer points to the legislative history of the Copyright Act, which explicitly states that these categories are “‘illustrative and not limitative,’ and … do not necessarily exhaust the scope of ‘original works of authorship’ that the bill is intended to protect.” ¹⁵

Although it seems clearly incorrect for the Office to assert that it is precluded by statute from registering subject matter that does not fall within one of the enumerated categories, one can understand why the Office, as a matter of policy, might adopt this position. In 1980, the Supreme Court faced a similar question in Diamond v. Chakrabarty, when it was called upon to decide whether genetically engineered living organisms could be patented. ¹⁶ Several dissenting Justices argued that the Court should not take the lead in expanding the scope of patentable subject matter to encompass new inventions arising out of biotechnology, but that instead the decision to expand patent law to biotechnology should be left to Congress. The majority rejected this approach, however, and held that Congress intended patent law to be dynamic and capable of expanding to accommodate important new developments in technology such as biotech. In retrospect, many people point to Chakrabarty as a pivotal event that helped encourage investment in biotechnology by reassuring investors that patent protection would be available for resulting innovations. Significantly, Congress has never explicitly authorized the extension of patents to new technologies like biotechnology, but instead has permitted the courts to oversee the expansion of patentable subject matter to accommodate new technologies.

Historically, copyright law has also expanded to encompass new technologies without explicit congressional action. Software is a good example. Today the availability of copyright protection for computer programs is well established, but the Copyright Statute has never been explicitly amended to provide for the copyrightability of software. Instead, it has largely been the courts that have overseen the expansion of copyright to encompass even highly functional software products written in object code and fixed in media that can only be interpreted by a computer.

The legislative history of the Copyright Statute indicates that Congress intended to allow for its expansion to accommodate new technologies. The law has a long and well-established tradition of using analogy as a primary mechanism for adapting to the development of new technologies that, in view of the overriding policy objectives underlying copyright law, warrant copyright protection. For example, Professor Nimmer has observed that “[a]s to new forms of creative expression that may emerge in the future as a result of scientific discoveries or technological developments … [i]f such a new form is sufficiently analogous to the kinds of works that are expressly protected in the eight categories, it will be regarded as falling within ‘the present congressional intent,’ even though the similarity is only by analogy.” ¹⁷ Similarly, in his treatise, Professor Goldstein correctly observes that “the question will sometimes arise whether a new form of authorship, not expressly mentioned in the Act, is entitled to protection. The most practical and principled approach to this question is to reason by analogy to works expressly listed in section 102. … new forms of works should be protected if they are similar to those listed and not protected if they are dissimilar.” ¹⁸

As one of us has pointed out in a previous article, the analogy between software and engineered genetic code is striking and becoming ever more so as the two fields evolve and converge. ¹⁹ A 2010 article, co-authored by a group of biologists, computer scientists, and computational biologists, makes this point nicely:

Given that counting genes in the genome is such a large-scale computational endeavor and that genes fundamentally deal with information processing, the lexicon of computer science naturally has been increasingly applied to describing them. In particular, people in the computational biology community have used the description of a formal language to describe the structure of genes in very much the same way that grammars are used to describe computer programs—with a precise syntax of upstream regulation, exons, and introns. Moreover, one metaphor that is increasingly popular for describing genes is to think of them in terms of subroutines in a huge operating system (OS). That is, insofar as the nucleotides of the genome are put together into a code that is executed through the process of transcription and translation, the genome can be thought of as an operating system for a living being. Genes are then individual subroutines in this overall system that are repetitively called in the process of transcription. ²⁰

Nonetheless, the Denial rejected our arguments with respect to the close analogy between engineered genetic code and computer code. The Office's primary justification for its divergent treatment of genetic and computer code seems to be based on Congress's decision in 1980 to amend the Copyright Statute to include a definition for the term “computer program.” The Copyright Office appears to believe that in making this amendment, Congress intended to expand the scope of copyrightable subject matter to include computer programs. Thus, according to the Denial, computer programs are to be treated differently from other new forms of expression made possible by advances in technology, no matter how high the degree of analogy between computer programs and other new forms of expression, like engineered genetic code. In our view, the Office's position reflects a misunderstanding of the significance of the 1980 amendment.

The context of the amendment is significant in this regard. In 1980, Congress made two amendments to the Copyright Statute relating to computer programs. First, it completely rewrote Section 117, renaming it as “Limitations on Exclusive Rights: Computer Programs.” This new Section 117 created certain limitations on the enforceability of copyright on computer programs. Second, Congress introduced a definition of the term “computer programs” into Section 101 of the statute, presumably in an attempt to reduce ambiguity with respect to the proper interpretation of new Section 117. Significantly, Congress has never amended the Copyright Statute to explicitly extend copyright protection to computer programs. The most plausible explanation for this is that Congress believed that computer programs were already copyrightable under the Copyright Statute. Indeed, this was the conclusion of the CONTU Report, which was written by a panel of experts established by Congress in the 1970s to study and make recommendations as to how the copyright law should respond to the increasing commercial significance of computer programs. ²¹

If Congress had not believed that the copyright statute already encompassed computer programs, it could, and presumably would, have amended the statute to explicitly cover computer programs. For example, as originally enacted in 1976, Section 102(b) of the statute only enumerated seven illustrative categories of copyrightable subject matter. ²² When Congress decided in 1990 to include “architectural works” within the realm of copyrightable subject matter, it explicitly amended Section 102(b) by adding “architectural works” as an eighth enumerated category. ²³ Thus, while the Copyright Office treats the 1980 amendments as evidence that Congress wished to limit the expansion of copyright to computer programs, a better interpretation is that the amendments evidence the fact that Congress viewed 102(b) as providing a broad and expansive, non-limiting definition of copyrightable subject matter that, as a general matter, encompasses new forms of expression made possible by advances in technology. The Appeal provides no explanation as to why computer code should be treated any differently than genetic code. In fact, most of the important justifications for extending copyright to computer programs cited in the CONTU Report apply equally to engineered genetic sequences. ²⁴

It bears noting that the Office's current stance regarding the registration of engineered genetic sequences is entirely inconsistent with the approach it took with respect to software in the 1960s. People began attempting to register computer programs in the early 1960s, and the Copyright Office initially expressed “profound doubts” as to whether computer programs qualified as copyrightable subject matter. ²⁵ Nonetheless, in 1964 the Copyright Office began to permit registration of computer programs under its “rule of doubt[], leaving the ultimate question of copyrightability to the courts.” ²⁶

Furthermore, the Denial provides no convincing rationale for the Office's decision to treat computer code, but not genetic code, as a form of literary work. Admittedly, the characterization of computer programs as “literary works” is quite a stretch, ²⁷ but the stretch occurred decades ago, when the courts and Copyright Office began treating highly functional computer programs, written in object code only interpretable by a computer, as literary works. Given the growing analogy between engineered genetic code and computer code, there is little if any additional stretch in going from computer code to engineered genetic sequences.

The Denial argues that the use of a sequence of letters to represent the Prancer DNA sequence does not constitute “literary authorship, but rather a form of notating the biological sequence,” because “[e]very gene sequence is represented by some sequence of these four letters to identify the sequence of the chemical compounds that these letters represent.” But once again, the Office fails to explain how this is any different than the case of software. It is well established that even a computer program written in object code, i.e., as a string of zeros and ones interpretable only by a computer, is copyrightable subject matter. But this string of zeroes and ones is nothing more than “notation” (to use the language of the Denial) that identifies, in the case of an optical storage device such as a CD-ROM or DVD, a sequence of deformities on the surface of a circular disc, or, in the case of a magnetic storage device such as a floppy disk or hard drive, a pattern of magnetization on a magnetically coated surface. The fact that computer code and genetic code “notate” the physical attributes of a medium used to convey information to a machine or device does not render the underlying code uncopyrightable per se.

The Office's inability to perform a search for similar sequences

The Denial asserts that the Copyright Office lacks the ability to perform a search of naturally occurring DNA sequences in order to determine whether a sequence submitted for registration is derived from a naturally occurring sequence. This is essentially a policy, as opposed to legal, basis for refusing registration. What is not clear is why the Office would feel it is necessary to perform such a search. As a general matter, the Copyright Office does not perform searches to determine if subject matter that has been submitted for registration is identical or similar to pre-existing works. ²⁸ Moreover, the Office presumably lacked the ability to perform searches with respect to computer programs it registered in the 1960s under the rule of doubt. The Denial provides no explanation as to why there is a unique need to perform a search with respect to an engineered genetic sequence submitted for registration.

Furthermore, if the Copyright Office deems it necessary or desirable to perform searches on engineered genetic sequences submitted for registration, then it would seem incumbent upon the Office to attain that capability, rather than denying registration to copyright-eligible subject matter. The Patent and Trademark Office (PTO) has not always been equipped to search for prior art in the realm of DNA sequences (or software for that matter), but as inventors began applying for patents on these technologies, the PTO responded by developing the capabilities. ²⁹

A group of copyright experts recently recommended that the Copyright Office consider outsourcing some registration responsibilities to third parties as a means of addressing certain limitations in the capabilities of the Office. ³⁰ Registration of engineered genetic sequences could be an area where such outsourcing would be particularly useful, particularly if the Office concludes that it is important to perform searches prior to registration. For example, the Copyright Office could delegate DNA sequence searches to the PTO, or some other agency with more expertise in searching and examining genetic sequences, or even to a private third party. ³¹

Overlapping copyright and patent protection

The Denial goes on to conclude that the “inability of the Office to independently discern new creative authorship [i.e., to perform a prior art search] suggests that a claim in a DNA sequence may be far better suited for the realm of patent, where a heightened standard of novelty, nonobviousness, and an examination of prior art would be considered, rather than the originality standard of copyright.” It goes on to assert that the fact that engineered DNA sequences are patent eligible “provides reason to question whether synthetic or cDNA sequences are proper subject matter for copyright,” pointing to a “potential overlap between copyright and patent protection.” But there is absolutely no reason why both forms of protection cannot overlap with respect to the same subject matter. The Federal Circuit recently rejected any suggestion that the availability of patent protection for software should render it ineligible for copyright protection, emphasizing that the Supreme Court has made clear that “[n]either the Copyright Statute nor any other says that because a thing is patentable it may not be copyrighted.” ³²

The Denial also makes much of the fact that “Congress has chosen to provide delineated patent protection for certain biological processes, such as plant patents for newly invented strains of asexually reproducing plants.” But the suggestion that the availability of plant patents somehow negates the availability of alternate forms of intellectual property protection for biotechnology was effectively refuted by the Supreme Court in J.E.M. Ag Supply, Inc. v. Pioneer Hi-Bred Int'l. ³³

The Office's assertion that engineered sequences are uncopyrightable under 17 U.S.C. Section 102(b)

The Office goes on to assert that genetic code is uncopyrightable under Section 102(b) of the copyright statute, which prohibits the extension of copyright to “any idea, procedure, process, system, method of operation, concept, principle, or discovery.” According to the Denial, “The Prancer DNA Sequence is genetic formula for a biological system. It does not describe, explain, or illustrate anything except the genetic markers that comprise this biological organism. Therefore, there is no copyrightable expression, but rather the claim simply records the formula for this biological system or process.”

At this point the Office seems to have entirely lost sight of the nature of engineered DNA. The reference to “genetic markers” and “this biological organism” make no sense—an engineered genetic sequence does not constitute a genetic marker, nor is it a biological organism. The characterization of the sequence as a “genetic formula for a biological system” is likewise difficult for a biologist to interpret, and appears to reflect a lack of understanding.

In any event, the analogous argument that copyright for software is precluded by Section 102(b) has been raised from time to time, and repeatedly rejected by the courts. For example, in Oracle v. Google, the court held that “components of the program that can be characterized as a “method of operation” may nevertheless be copyrightable. The Oracle court also pointed out that “classifying a work as a ‘system’ does not preclude copyright for the particular expression of that system.”

The functionality and alleged artistic deficiencies of engineered sequences

The letter acknowledges that “[i]n the case of a synthetic gene, the specific sequence of nucleotides is the result of some person's choices,” but goes on to conclude that these choices are not sufficient to render the resulting sequence copyrightable because “those choices are not made for the purpose of artistic expression. … Properly understood, the nucleotide sequence of a synthetic gene, inasmuch as it could be conceived as a form of expression at all, is a form of expression dictated solely by functional considerations.”

But it is incorrect to suggest that copyright protection is limited to “artistic” works, and denied to functional works. It is well established that even highly functional software can be copyrighted, regardless of whether it incorporates any artistic element that could be discerned by the Copyright Office. Indeed, in Oracle, the court recognized that “computer programs are by definition functional,” and that Congress and judicial precedent firmly establish that computer programs are copyrightable, “despite their utilitarian or functional purpose.” ³⁴ Of course, one might argue that the author of even the most functional computer program might incorporate elements that would be perceived as aesthetically pleasing by other software designers, but the same could be said for the design of a synthetic genetic sequence. As an aside, there are examples of non-functional synthetic DNA sequences that would easily satisfy even the “artistic” creativity threshold proposed by the Office, including literary quotations introduced into a synthetic M. genitalium genome by Craig Venter's group and a version of George Church's book, Regenesis—How Synthetic Biology Will Reinvent Nature and Ourselves, encoded in DNA. ³⁵

A bias against biological creativity and works based on nature

The Denial characterizes engineered DNA as the “result of … biological techniques,” and asserts that “[b]iological creations do not fit within any of the existing categories of authorship.” It goes on to state that it is unclear whether a DNA sequence can ever be “entirely human-created,” noting that “[s]ince the operation of DNA is dictated by laws of biology, at least some aspects of DNA sequences are controlled by those laws.” Without further explanation from the Office, it is hard to know exactly what to make of such assertions. To say that the operation of DNA is dictated by laws of biology seems comparable to saying that the operation of software is dictated by the laws of physics, which is of course true, but does not preclude copyrightability.

The Copyright Office seems to be expressing a bias against the idea of copyright extending into the realm of biology. Professor Kayton himself confessed to the same sort of initial prejudice when he was first confronted with the idea of copyright for engineered genetic sequences. His prejudice was only overcome after he grappled with the issue for some time and was forced to acknowledge that logic dictated that copyright protection for computer code imputes copyright protection for genetic code.

Similarly, in response to our argument that DNA is analogous to a computer program because it is a set of instructions directed towards a biological machine, the Office asserts that “[w]hile an organism may be analogized to a machine, it is clearly not one, and as a result falls outside of the category enumerated in the statute”; i.e., it is not a literary work. The Copyright Statute specifies that copyright protection is available for “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.” ³⁶ So it appears that the Office is arguing that the determinative difference between computer code and genetic code is that computer code is directed towards a computer, which is a “machine or device” as explicitly provided for in the statute, while genetic code is directed towards a biological organism or system which falls outside the bounds of “machine or device.”

Once again, this seems to be a clear case of discrimination against biological-based innovation. The Copyright Statute explicitly defines the terms “machine or device” as broadly encompassing machines and devices “now known or later developed.” ³⁷ The Merriam-Webster online dictionary defines a “device” as “an object, machine, or piece of equipment that has been made for some special purpose.” A “machine” is defined as “a piece of equipment with moving parts that does work when it is given power from electricity, gasoline, etc.” Both definitions, and particularly the definition of device, would appear to encompass a biological organism programed to perform some useful function, especially given that the Copyright Statute explicitly anticipates expansion of the definition of device and machine to keep up with ongoing advances in technology.

In fact, synthetic biologists often refer to engineered biological organisms as “machines,” and at times even as “computers.” For example, researchers at Stanford University recently reported using engineered DNA to create inside living cells a “biological device [that] behaves like a transistor, one of the tiny switches that are etched on to microchips in the billions to perform computer calculations.” ³⁸ In an article published in the scientific journal Science, these researchers explained “how their biological transistors could be connected together inside living cells to perform computing jobs such as controlling how genes are expressed in an organism.”

The Denial goes on to state that “the extent to which any genetic sequence can truly be entirely human-created is unclear, particularly with respect to copyright creativity,” because “to the extent that any sequence involves the isolation of any naturally-occurring sequence, or a derivation thereof, it may not be deemed created as opposed to discovered.” In essence, the Office seems to be suggesting that the fact that engineered genetic sequences are, at some level, based upon a modification or reconfiguration of genetic information ultimately derived from naturally occurring sequences somehow renders engineered genetic sequences uncopyrightable products of nature. But there is no prohibition against copyright protection for an expressive work based upon nature. The photography of Ansel Adams, capturing the natural beauty of Yosemite Valley, for example, is not precluded from the protection of copyright.

In Proline Concrete Tools, Inc. v. Dennis, a district court recently chastised the Office for making a similar mistake when it denied registration for “rock and stone sculptures … used to make decorative concrete stamps.” ³⁹ The decision to deny was based on the Office's conclusion that the sculptures were “slavish copies of un-copyrightable objects and, as such, do not contain a sufficient amount of original authorship to support copyright claims.” The court overturned the Office's decision, however, characterizing it as the result of “a clear misunderstanding on the part of the Register that seemed “to result from its confusion over how [the registrant] creates the sculptures and what exactly it seeks to copyright.” ⁴⁰

Similarly, in the case of Prancer, the Office seems to misapprehend the nature of genetic engineering and has conflated the copyrightability of a naturally occurring genetic sequence, which no one is arguing should be eligible for copyright, with the copyrightability of an engineered genetic sequence. Like the stone sculptures in Proline Concrete Tools, the Prancer sequence is not a slavish copy of a naturally occurring genetic sequence, and registration of its sequence would in no way threaten the public's access to any naturally occurring sequence. In large part, the denial of registration appears to be based on the Office's unwillingness to recognize the fundamental distinction between naturally occurring sequences and a sequence that diverges significantly from any naturally occurring sequence from which it may have been derived.