A(rtificial) I(ntelligence) Created This, How Can I Protect It? BY ALEXANDER C. KACZMAREK, HARTER SECREST & EMERY, LLP 14 | September/October 2023 | BAEC Bulletin Cyber Law Artificial intelligence
from understanding AI’s black box and its potential impact on the legal system and practice of law as a whole, a more useful and immediate question is: How does copyright law treat the seemingly new content that generative AI “creates”? Copyright law protects original works of human authorship, including text, images, and music. In Burrow-Giles Lithographic Co. v. Sarony, the leading case on authorship, the Supreme Court repeatedly referred to “authors” as human 1 . The Ninth Circuit has held that authorship excludes “non-human spiritual beings” 2 and animals. In its current edition of the Compendium, the Copyright Office states, “to qualify as a work of ‘authorship’ 3 a work must be created by a human being” and that it “will not register works produced by a machine or mere mechanical process that operates randomly or automatically without any creative input or intervention from a human author.” 4 Together, these sources have been synthesized by the Copyright Office to evidence the human authorship requirement as well-established and, accordingly, the Copyright Office excludes generative AI content from protection because of its non-human authorship. But what about the Copyright Office’s own stated exception for “works produced by a machine” with the “creative input or intervention from a human author?” 5 To enable the generative AI program to produce content, a human enters an idea for the content into a selected program, which determines the style, medium, and form of the output. The Copyright Office’s position is that this level of creative control isn’t enough. Instead, it likens the act of entering a prompt into a generative AI program as akin to providing instruction to a commissioned artist; where the final work is not created by the human providing instruction, but the recipient of the instruction. 6 Thus, the contribution by the human providing instruction is viewed as the unprotectable idea, not the protectable expression of that idea. So, where do we go from here? Is it possible to obtain
isn’t new, but what it can accomplish is. From science fiction novels to a search engine’s suggestion bar, artificial intelligence (“AI”) has developed from mere fantasy to a familiar part
of everyday life. The emergence
of “generative” AI pushes past automation and assistance toward creation and innovation. But, pulling back the curtain, what is that creation and innovation really made of, and can that creation be protected under copyright law? To answer the first half of our question, let’s turn to generative AI itself (the remainder of this paragraph was drafted by a generative AI program):
Generative AI operates by employing deep learning neural networks. These models are trained on vast datasets to learn patterns and relationships within text, images, or other data types. When prompted with input, the AI generates coherent and contextually relevant output by predicting the most likely next elements based on its training. It relies on probability distributions to generate diverse and creative content, making it useful for tasks like text generation, language translation, image synthesis, and more. The AI’s ability to generate human-like content stems from its capacity to capture intricate linguistic and contextual nuances from its training data.
The current version of generative AI is just the beginning of what it could and will be. Even when limiting the question to generative AI’s impact on a specific area of practice, there are countless questions and issues that arise. So, shifting our focus
Powered by FlippingBook