Originality in AI Art: A New Kind of Author

The question of what makes art original has occupied artists, critics, and philosophers for generations. Generative artificial intelligence has made that question more urgent than ever. When a machine produces an image, who can claim to be its author? And can the result be considered original? For artists working at the intersection of AI and digital creativity, these are not abstract debates. They shape the way work is made, presented, bought, and understood. They also influence the way audiences respond, the way galleries display the work, and the way the law treats it.

What Does Originality Mean in AI Art?

AI generates images by analysing vast amounts of existing work and mimicking patterns. This process immediately raises questions about originality and authorship. If the machine draws on millions of pictures that already exist, how can its output ever be genuinely new? The question is not academic. It sits at the heart of a growing controversy within the art world, where some celebrate the technology and others fear its effects.

The legal definition offers one way in. Originality has been defined as independent creation plus a minimal degree of creativity. In human art, that test is usually satisfied without much difficulty. The artist creates the work independently and makes enough creative choices to cross the threshold. When AI is involved, the test becomes harder to apply.

Some researchers argue that AI-generated artworks can be original and that this should not be a cause for concern for either artists or lawyers. Under this view, the human artist who guides the machine through carefully chosen prompts, selects the strongest candidates, and edits the final output may still satisfy the independent creation requirement.

The role of artificial intelligence in design also affects the concept of originality in less obvious ways. The machine is not simply a new brush. It is a system trained on the work of others, and that fact lingers over every image it produces.

Who Is the Author When a Machine Creates?

The new kind of author is hard to place. Is it the person who writes the prompt? The engineers who built the model? The machine itself? Copyright law has traditionally reserved the title of author for human beings.

A philosophical analysis published in the Houston Law Review frames originality as independent creation plus a minimal degree of creativity. By extending copyright protections to art generated by AI, the legal system would effectively treat the machine as a kind of author. That step carries significant consequences for how creativity is understood.

Others argue the opposite. A note from Ave Maria School of Law, published in August 2024, states that AI-generated art is not copyrightable because it does not satisfy the originality element of a copyrightable work. If the machine did the creating, then the human contribution may not be enough to meet the legal standard.

The comparative study of copyright laws, which draws on the work of artists such as Refik Anadol, suggests that the final answer is far from settled. Different legal systems may reach different conclusions about the originality of AI-generated artworks. What copyright laws can learn from these works remains an open question.

The Limits of Machine Creativity

Generative AI is often described as a remixing engine. For good reason: to become truly creative, generative AI must go beyond remixing data and begin to mirror human emotion, intuition, and imagination. That is a demanding standard, and current systems rarely meet it. They can imitate the surface of a style, but they do not understand why that style matters, what it expresses, or how it connects with the viewer.

There is also the question of the data itself. A machine cannot invent a visual language from nothing. It can only combine and transform what it has already seen. This is why the output can feel derivative even when it is technically impressive. The imitation is skilful, but it rarely surprises in the way that genuine invention does.

Yet research by Zhou and colleagues suggests that the story is more complicated. AI-assisted artists who successfully explore more novel ideas, regardless of their prior originality, may produce artworks that their peers recognise as innovative. The machine can push the human artist to think differently, and the resulting work can be original in a meaningful sense.

The key may be the interaction itself. The human supplies the direction and the meaning; the machine supplies the speed, scale, and unexpected combinations. Originality in AI art emerges from that collaboration, not from the algorithm working alone.

The Power of Human Connection

There is a strong argument that the trouble with AI art is not just a lack of originality. In a May 2025 opinion piece for The Guardian, Eric Reinhart wrote that when artwork is invented by a machine, it loses its most important power: to help people connect. In an already lonely era, that loss carries real weight.

The same article noted how ChatGPT users have been imitating the style of Hayao Miyazaki, the celebrated animator behind films such as The Boy and the Heron. The imitation may look convincing, but it carries none of the intent, care, and personal history of the original artist. What remains is surface, not substance.

This argument reframes the debate. Originality in AI art is not only about whether the image is new. It is about whether the image means something. A machine can generate a picture, but it cannot generate a shared experience between artist and audience. It cannot know what a viewer will feel, and it cannot intend to move them.

For human artists, this is encouraging. The capacity to connect remains firmly on the human side of the collaboration. No algorithm can replace the understanding that passes between the person who makes the work and the person who receives it.

A New Kind of Author

What does the future hold for the author in the age of AI? The most likely answer is not the disappearance of the author but the arrival of a new kind of author: the AI-assisted artist.

This artist uses the machine as a collaborator rather than a replacement. They write the prompts, curate the results, impose meaning on the output, and stand behind the finished work. The originality lies in those choices, not in the algorithm alone.

For an artist working with AI-generated and pixel-art digital artwork, the opportunities are real. Online galleries such as ArtPal and Fine Art America allow artists to sell original pieces, and buyers are increasingly curious about the process behind the work. The story of creation has become part of the product itself.

The new author also carries a responsibility. They must be honest about the role of the machine while insisting on the value of the human hand. Originality in AI art is not an illusion, but it is a shared achievement. The machine may generate the image, but the artist generates the meaning, and that is where true authorship begins.

Frequently Asked Questions

Here are answers to some of the most common questions about originality, authorship, and the legal status of AI-generated art.

Can AI art be considered original?

The debate over originality in AI art is far from settled. Some researchers argue that AI-generated artworks can be original, while others claim that the machine’s reliance on existing images disqualifies them. The legal definition of originality as independent creation plus a minimal degree of creativity offers a framework, but applying it to machine-made work is difficult. Much depends on how much human involvement is required, and different jurisdictions may reach different answers.

How is AI art made?

AI generates images by analysing vast amounts of existing work and mimicking patterns. The system learns from a large dataset of images and uses that knowledge to create new pictures in response to text prompts. The quality of the output depends heavily on the training data and on the skill of the human operator in guiding the machine towards a desired result.

Can I legally use AI-generated art?

The legal position on AI-generated art is still developing. Some legal commentary holds that AI-generated art is not copyrightable because it does not satisfy the originality element of a copyrightable work. This means your rights over a particular AI-generated image may be limited. If you intend to sell or licence AI art, you should check the current rules in your jurisdiction and keep clear records of your creative input.

Is AI replacing human artists?

The research points to a more collaborative relationship rather than a takeover. AI-assisted artists who explore novel ideas may produce work that their peers recognise as innovative. The machine can generate images quickly, but it lacks the human emotion, intuition, and imagination required for deeper creativity. Most evidence suggests that AI is becoming a tool for artists, not a replacement for them.

What is the 30% rule in AI?

The available research does not support a recognised 30% rule for AI-generated art. Claims about such a threshold appear in online discussions, but no consistent legal or industry standard is established in the sources reviewed. If you are making commercial decisions based on a specific percentage, you should verify the current position with the relevant copyright office or a qualified legal professional.