AI and the Artist: Creativity, Competition and the Law

Creative

AI can be a powerful tool for artists. Apps like Chord AI identify chords in songs, helping musicians translate inspiration into new work. Playlist AI creates playlists from photos, videos, text or listening history, making discovery easier for listeners and creators.

However, other AI tools learn from artists’ work and reproduce it when prompted. This makes it harder for artists to differentiate their style. Greg Rakowski recently described how AI software generated images almost indistinguishable from his own art, without his consent. He has joined a lawsuit against Stability AI, Midjourney and DeviantArt alleging copyright infringement and unlawful competition. The legal implications of this will be explored below.

Economic

Some artists have begun tagging their work with labels such as “noAI” to prevent it from being scraped and fed into training datasets. The concern is not only creative imitation but economic survival. As Greg explained in The Journal podcast, artists are commissioned because of their unique style. If AI can replicate that style for free, it becomes a direct competitor.

Art-generating software threatens the livelihoods of artists whose income depends on distinctiveness. When a machine can produce a convincing imitation in seconds, the value of human originality becomes harder to protect.

Legal

Unfair competition is one of the claims raised in the artists’ lawsuit. The central allegation is that Stable Diffusion uses artists’ work without consent. The process involves feeding the AI vast numbers of images. When prompted to create art in the style of Picasso, for example, the system can do so because it has learned what Picasso’s work looks like. It can then generate a Picasso‑like piece without permission from Picasso’s estate.

This raises the question: is this legally wrong?

AI creativity mirrors human creativity in some ways. DALL‑E, for instance, is trained on existing works, internalises them and produces new pieces when prompted. Human artists also absorb inspiration from galleries, music and other creators. They reflect on what they have seen and produce new work influenced by it. Both the artist and the AI generate new creations inspired by existing material.

Under the Infopaq standard, a work must be the author’s own intellectual creation to qualify for copyright protection. At first glance, AI output appears to follow a similar process.

However, as a friend researching AI law recently pointed out, the standard for a machine cannot be the same as the standard for a human. The Human Artistry Campaign’s newly published principles argue that copyright should protect the unique value of human creativity. Although AI and humans may operate similarly in theory, in practice a machine can copy at a speed and scale that humans cannot. The resulting work may not be merely inspired but effectively a copy.

Juls, the Grammy award winning producer, put it simply: there is a human touch in art that a machine cannot replicate. I would add that there is a human touch that even another human cannot replicate.

Conclusion

AI will undoubtedly transform the artistic landscape. Some tools will be invaluable for artists and consumers, especially those that support creativity and discovery. Yet as AI evolves beyond a tool and becomes a force in its own right, it will challenge the status quo and make it harder for artists to earn a living.

Although deep learning appears to align with existing copyright principles, it remains to be seen whether the law will adapt and impose a different threshold for AI. The future of art will depend on how society chooses to balance innovation with the protection of human creativity.