The Paradox of Originality: How Early AI Art Challenged Copyright & Authorship

You’ve seen the headlines: AI-generated art winning competitions, blockbuster movie posters created from a text prompt, and the inevitable, complex lawsuits that follow. The debate over artificial intelligence, creativity, and ownership feels incredibly modern, a product of our hyper-connected 21st-century world.

But what if I told you that today’s controversies are just echoes of a conversation that started over a cup of coffee and a clunky computer in the 1990s?

The questions we're asking now—Who is the artist? Can a machine create? Who owns the final image?—aren't new. They were first whispered in university labs and art galleries decades ago, when the first truly autonomous AI art systems began to blur the lines between programmer and painter, code and canvas. This is the story of that first digital identity crisis and how it laid the groundwork for the world we live in today.

Before "Prompt Engineer" Was a Job Title: AI Art in the 90s

Forget the photorealistic images you see today. AI art in the 1990s and early 2000s looked very different. It wasn't about a massive neural network learning from the entire internet. It was about something far more intimate: a single artist teaching a single computer how to create.

The undisputed pioneer of this era was Harold Cohen, an artist and computer scientist who spent decades developing a program called AARON. Think of AARON not as a magic black box, but as Cohen’s lifelong creative partner. He didn't just write code; he painstakingly taught AARON the rules of art: how to compose a scene, how to draw figures, how to choose and apply color.

AARON wasn’t a tool in the way Photoshop is a tool. It was a generative system. Cohen would turn it on, and AARON would begin to draw and paint on its own, producing original works that Cohen himself could not have predicted. Each piece was a genuine surprise, a true collaboration between human and machine. This novel approach was a masterclass in , pushing the boundaries of what a digital tool could be.

This unique partnership immediately sparked a philosophical and legal firestorm. If Harold Cohen didn’t know exactly what AARON was going to paint, could he truly be considered the sole author?

The Core Dilemma: Who Is the "Author"?

Copyright law, at its core, is built on a very human idea: a person has a flash of inspiration and fixes it in a tangible form (a book, a song, a painting). This "human authorship" is the bedrock of intellectual property.

AARON crashed head-first into this principle. It raised a paradox that lawyers and artists are still trying to solve:

• The Argument for Human Authorship: Team Cohen argued that he was the undeniable author. He wrote every line of AARON’s code, imbued it with his artistic knowledge, and curated its output. The AI, they claimed, was simply an incredibly sophisticated tool, like a paintbrush that happened to run on electricity. The creative spark originated with him.
• The Argument for a New Paradigm: Others weren't so sure. AARON made decisions. It chose colors and arranged shapes in ways that were emergent and unpredictable. Did this emergent behavior grant the program a sliver of authorship? Since a machine couldn't legally be an author, did that mean the art was in the public domain?

This wasn't just a dorm-room debate; it struck at the heart of what we define as creativity. Is art about the final, unpredictable product, or the initial, intentional act of creation? Exploring these early questions reveals so many of the and how our understanding has evolved.

A Legal System Playing Catch-Up

While artists and philosophers debated, the legal system had to find a practical answer. In the United States, the Copyright Office held a firm line: a work must be created by a human being to be copyrightable. A photograph of a monkey couldn't be copyrighted. A mural painted by an elephant couldn't be copyrighted. And an image generated solely by a machine fell into the same category.

The most useful analogy at the time was photography. A camera is a machine that captures an image, yet the photographer is granted copyright. Why? Because the photographer makes a series of creative choices: composition, lighting, focus, and the decisive moment to click the shutter.

The legal question for AARON became: Is writing the code for an art-generating AI analogous to taking a photograph? The consensus leaned toward "yes." Harold Cohen's creative input was in designing the system itself. His art wasn't just the final image; it was the elegant, complex set of rules that made the image possible. Therefore, he could claim authorship over AARON's output because it was the direct result of his creative expression as a programmer and artist.

Why the 90s Debate Still Echoes Today

The framework established around AARON—that the human behind the machine is the author—held steady for decades. But today's generative AI models have thrown a wrench into that tidy conclusion.

The questions are the same, but the scale and complexity have magnified exponentially.

• Then: The debate was about one artist teaching one AI a specific style. AARON's knowledge came entirely from Harold Cohen.
• Now: The debate is about massive AI models trained on billions of images scraped from the internet, created by millions of human artists, most of whom never gave their consent.

The core question of "authorship" is no longer just about the user and the machine. It now involves the creators of the original training data, the company that built the AI model, and the user who writes the prompt. The paradox of originality is more complex than ever, but its roots are firmly planted in the pioneering work of artists like Harold Cohen who first dared to ask, "What happens when the paintbrush starts to think?"

Frequently Asked Questions About Early AI Art & Copyright

What was the first truly "AI" artwork?

While computer-assisted art dates back to the 1960s, Harold Cohen's AARON is widely considered the most significant and long-running early example of "AI art" because of its autonomy. It could create original images from a complex internal set of rules without step-by-step human guidance for each piece.

Could you copyright art made by AI in the 1990s?

This is the key distinction: you could not copyright a work authored by an AI. However, the U.S. Copyright Office's stance allowed for humans to claim copyright over works created using an AI as a tool, as long as there was sufficient human creative input in the process. For Harold Cohen, that creative input was the decades he spent writing and refining AARON's code.

Who was Harold Cohen?

Harold Cohen (1928-2016) was an accomplished British painter who became a pioneering computer artist. After becoming a professor at the University of California, San Diego, he began working on AARON in the early 1970s. He dedicated the rest of his life to exploring the potential of artificial intelligence as a partner in the creative process.

How is the old debate different from today's AI copyright issues?

The primary difference is the data. Early systems like AARON were "closed-box." They created art based only on the rules their creator gave them. Today's AI models are trained on vast datasets of existing human-made art, which introduces critical new legal questions about copyright infringement, fair use, and derivative works that simply weren't part of the conversation in the 90s.

The Unfinished Conversation

Understanding the history of AI and art isn't just an academic exercise. It provides a crucial framework for navigating the challenges and opportunities we face today. The questions about authorship, creativity, and the role of technology in art are not going away. They are simply evolving.

By looking back at the pioneers who first wrestled with these ideas, we can better appreciate the complexity of the modern AI landscape and contribute to a conversation that is, in reality, decades in the making. The next time you see a stunning AI-generated image, remember the story of AARON—a simple program that asked a very complicated question, forever changing the relationship between the artist and their tools.

If you're inspired to see what modern creators are building, explore this gallery of that are continuing this legacy of innovation.