140 THE US-ISRAEL | Legal Review 2025/26 a machine produces a photorealistic image, a compelling text, or a symphonic movement in response to a short “prompt,” it challenges the notion that creativity is a unique human attribute. The Changing Nature of the Creative Process To understand the legal crisis, one must also understand the technological metamorphosis. Historically, technology acted as a “passive instrument.” The camera, once feared as the death of painting, still required the photographer to choose the angle, the lighting, the focal point, and the decisive moment of capture. The computer allowed for digital editing and sophisticated manipulation, but the artist remained the primary conductor, making thousands of micro-decisions—from the pressure of a digital pen to the opacity of a layer— that shaped the final work. From Tools to Autonomous Agents We are witnessing a transition through three distinct phases that redefine the human-machine hierarchy and, by extension, the legal standing of the resulting output: AI as a Tool: Traditional software where the human maintains full, granular control. Think of Photoshop filters or spell-checkers; these are digital extensions of the human hand, requiring active direction at every stage. In this phase, the machine’s role is purely subservient to the human’s “intellectual labor.” AI as a Collaborator: Systems that offer suggestions, co-write code, or provide variations on a theme. Here, the machine begins to show a form of “proactive” assistance, such as an autocomplete function that anticipates a writer’s intent or a generative design tool that offers multiple architectural iterations based on fixed parameters. The legal challenge here begins to simmer: if the machine suggests the “hook” of a pop song, who owns the melody? AI as a Generative Agent: Systems that produce highfidelity outputs based on abstract, natural language instructions (prompts). In this phase, the human retreats from the “execution” of the work to the “curation” and “direction” of the output. The AI makes the expressive choices—the specific hues of a sunset, the cadence of a poem, or the harmonies in a symphony. In this third phase, the “labor” of creation—the literal brushstrokes, the painstaking mixing of colors, or the arduous drafting of sentences—is automated. This automation strikes at the heart of the “Sweat of the Brow” doctrine (which, though largely rejected in favor of originality, still informs our intuition about value) and the “Modicum of Creativity” standard. We are moving from a world where we reward the “doing” to a world where we are asked to reward the “asking.” This shift threatens to commoditize creativity, reducing the value of human skill in favor of the efficiency of the prompt. The Dilution of Human Authorship As generative AI becomes more autonomous, the human’s role is becoming “thinned” or even “evaporated.” In the past, the artist was the “efficient cause” of the work. Today, when a user types “A sunset in the style of Van Gogh,” the AI interprets millions of statistical relationships between pixels to generate an image. The human did not choose the specific placement of a single brushstroke or the specific blending of hues; the machine’s latent space did. The question for the courts is whether “prompt engineering” is a creative act of authorship or merely a request for service, akin to a patron commissioning an artist. If I tell a painter to “paint me a portrait of my dog in a regal setting,” I am the source of the idea, but the painter is the author of the expression. GenAI blurs this line because the “painter” is now a set of weights and biases in a neural network. If the user provides no further input, they are acting as a curator rather than an author, a distinction that has profound implications for whether the output deserves the robust protections of copyright law. “The question for the courts is whether “prompt engineering” is a creative act of authorship or merely a request for service, akin to a patron commissioning an artist.”
RkJQdWJsaXNoZXIy MjgzNzA=