Human Creativity v. Machine Autonomy in Identifying Copyright Authors of Generative NFTs | Rothwell, Figg, Ernst & Manbeck, P.C.

0
82

Generative content is entering the moment in the form of non-fungible tokens (NFTs) powered by blockchain platforms. Continued investment and growth in the NFT market has provided a mainstream legal edge case for the use of digital technology to create intellectual property. Recently, a review board of the US Copyright Office has announced a decision to address whether the artificial intelligence “creativity machine” can meet the author’s statutory requirements for copyright purposes. look Decided dated February 14, 2022, available here. The Board acknowledged that the creativity machine does not meet the author’s statutory requirements. This is in line with Office’s position that authors must be human.

This article explores the broader meaning of the Board’s decision on generative content, including generative NFTs, and provides some practical guidance for those seeking to ensure copyright protection.

  1. Human copyright requirements for copyrighted works in the United States

Statistically, the copyright of a work “initially belongs to the author of the work”. 17 USC § 201 (a). “Author” is not explicitly defined, but the US Copyright Office takes the position that authors must be human.

Copyright law protects only “the creativity of the mind” and “the outcome of intellectual labor.” Trademark case, 100 US 82, 94 (1879). Because copyright law is limited to the “author’s original intellectual concept,” Office will refuse to file a claim if it determines that a person has not created the work. Burrow-Giles Lithographic Co.v. Sarony111 US 53, 58 (1884).

Summary of US Copyright Authority Practices, § 306. The Copyright Office further emphasizes that “the important question is whether the” work “is basically a human work and the use of a computer.” [or other device] Whether it is merely an auxiliary instrument, or whether the elements of the traditional author of the work (literature, art, musical expression, or elements such as selection, placement, etc.) were actually devised and implemented by machines rather than humans. ” Id.. , § 313.2.

US courts have taken a consistent view in interpreting the law as requiring human copyright. See, for example, Cmty. Creative nonviolence v.For leads490 US 730, 737 (1989) (“In principle, the author is the party who actually creates the work, that is, Man A person who transforms an idea into a fixed, concrete expression that is subject to copyright protection. “).The human author was focused on the case of Naruto vs. Slater, No. 15-CV-04324-WHO, 2016 WL 362231, * 3 (ND Cal. January 28, 2016), aff’d, 888 F.3d 418 (9thCir. 2018), also known as “Monkey Self-Shooting” There, photographer David J. Slater was in Indonesia to take pictures of wildlife. A 6-year-old Makaku named Naruto picked up the camera and took some pictures of himself. People for the Ethical Treatment of Animals (PETA) sought to qualify as Naruto’s author. However, the court ruled that Naruto is not a human and cannot be an author.

In light of the above case, the U.S. Copyright Office’s review board refuses to register 2D artwork “autonomously generated by a computer algorithm running on a machine” as a work for hire. I reaffirmed the human copyright requirements. A “creativity machine” consisting of deep artificial neural networks generated the above artwork.

The Supreme Court explained that the Supreme Court “has continued to clarify the relationship between the human mind and creative expression as a prerequisite for copyright protection.” Mather vs Stein347 US 201, 214 (1954) and Goldstein vs California, 412 US 546, 561 (1973). However, in particular, the Board’s decision is important: “Under what circumstances can humans be involved in the creation of machine-generated works to meet the statutory standards for copyright protection?” The problem remains unresolved.

  1. Meaning of “generative” in generative content

Generative content includes digital artwork such as images, videos, texts, and graphics created by computer programs that use some degree of autonomy. Artists typically instruct computer programs to use specific algorithms and, in some cases, create content within specific boundaries.

Many popular generative NFTs rely on some form of algorithmic support to randomly select variables within the artistic boundaries established by humans. For example, the CryptoPunks NFT collection contains stylized pixel art characters consisting of “algorithm-generated 24×24 pixel art images.” look Accessed https://www.larvalabs.com/cryptopunks, February 22, 2022. Similarly, the Bored Apes collection includes cartoon apes, each “uniquely and programmatically produced from over 170 possible traits such as expressions, hats, clothing, etc.” look Accessed https://boredapeyachtclub.com/#/home on February 22, 2022. On some platforms, such as ArtBlocks, users can select a collection, and in exchange for cryptocurrency payments, the platform will generate a new one with an algorithm on demand for a unique version of the content of the selected collection.

In contrast, certain advanced machine learning (ML) techniques, including generational modeling, give computer programs much more autonomy in the creation of digital content. Generated modeling uses an initial dataset as input, the program learns patterns and features about that data, and the program creates new data as output that may have been taken from the original dataset. There are several techniques for building and training generative model applications using neural networks, many of which today use Generative Adversarial Networks (GANs). GAN involves the use of two models. A generator that creates a new example and a classifier that classifies the new example as either from the original dataset (actual) or not (fake).

Generative models are used to create new information with various properties such as predicting the next word in a sequence, synthesizing training data, generating video predictions, and synthesizing voice or text. For example, GAN has become famous for being used to create “deepfake” content, such as image, audio, and video content that is visible but invisible.

  1. Is generative content subject to copyright?

The Board’s decision leaves unsolved the minimum level of human input required in the generated content creation process to meet the author’s statutory requirements. In essence, the question can be reduced to whether there is an identifiable relationship between the potentially creative elements provided by the human artist and the output from the computer program. A summary of U.S. copyright agency practices, §313.2 states, “Office is a work produced by a machine that operates randomly or automatically, or just a mechanical process, without the creative input or intervention of a human author. It provides some guidelines for “not registering”.

  • Reduce or increase the size of existing works.
  • Make changes to existing works as determined by manufacturing or material requirements.
  • Convert your work from analog to digital format, such as transferring videos from VHS to DVD.
  • Declick or reduce noise in existing recordings, or convert recordings from monaural to stereo sound.
  • Transpose the song from B major to C major.
  • Medical images produced by x-rays, ultrasound, magnetic resonance imaging, or other diagnostic equipment.
  • Claims based on a mechanical weaving process that randomly produces irregular shapes in the fabric without identifiable patterns.

In many generative NFT collections, there is almost certainly an identifiable relationship between the potentially creative elements provided by human artists and the output of the program. That is, humans program artistic boundaries regarding the composition of digital works, and artists use that program as a tool to create content within those artistic boundaries. In contrast, the creativity machines at issue in recent Board decisions use much more autonomous generated AI modeling technology to create digital art. With such advanced technology, the potential link between human input and machine output is undoubtedly diminished.

The exact legal boundaries between human creativity and machine autonomy continue to evolve with technology, and the Board said, “A US court examining whether artificial intelligence can be the author for copyright purposes. I don’t recognize it. ” It is important to note that the copyright applicant did not attempt to claim human copyright. Applicants seeking to ensure copyright protection for the generated content will name the human author and identify (preferably document) the contributions made by humans in defining the creative output produced by the computer program. It is recommended to do).

LEAVE A REPLY

Please enter your comment!
Please enter your name here