Artificial intelligence has moved from science fiction to a transformative force in contemporary art, challenging conventional views on both creativity and authorship. A.I. sets itself apart from earlier technological advancements like the printing press, photography and video as it actively participates in the creative process. Unlike previous advancements, which served primarily as mediums of reproduction or extension, A.I. introduces a dynamic feedback loop: it learns from the data it’s fed, responds to aesthetic input and evolves its outputs accordingly. This positions it as a semi-autonomous agent—one that can surprise its human collaborators and shift the direction of a work in real time. As such, the relationship between artist and machine becomes dialogic rather than directive.
Artists expanding the frontier of A.I.-generated art
Take Refik Anadol, whose breathtaking immersive installations leverage massive data collections and sophisticated A.I. algorithms—from weather data to architectural blueprints to brainwave activity—into emotionally resonant environments. His work demonstrates how A.I. can move beyond mere data processing to creatively interpret information and produce unique visual art. “For as long as I can remember, I have imagined data as more than just information—I have seen it as a living, breathing material, a pigment with infinite possibilities,” Anadol said during his TIME100 AI Impact Awards acceptance speech in February.
Then there’s teamLab, the collective earning recognition through their novel approach to machine learning. Their digital artworks respond in real-time to audience interactions and gestures, creating exhibitions—or better said, visual environments—that continuously evolve. A.I. enables this dynamic interaction by transforming the artwork and disrupting conventional roles for artists, viewers and the artwork itself.
Consider also Holly Herndon, an artist who composes music with an A.I. entity named “Spawn,” which has been trained using samples of her own voice as well as others to create musical responses. Her work expands traditional music boundaries and conceptions of authorship, using “Spawn” to create unique styles and sounds that combine human elements with machine-driven inputs. Herndon’s creative output shows how A.I. technology can help create entirely new paths for artistic expression. The artist collects her own data and trains her own model, co-creating with the machine at every single step of her artistic process.
In 2020, Krista Kim achieved notoriety with her project Mars House, the first digital NFT home. Rendered using A.I., the project captivated the imagination of the public as a space for wellness and tranquility, serving as a blueprint for real-world urban regeneration. Her work may sound utopian, but in professional circles has proved effective in querying the physical limitations traditionally associated with art and architecture. Mars House has since been cited in architectural think tanks and academic publications exploring digital placemaking and wellness brand initiatives aiming to develop immersive, meditative environments for future urban spaces. Her concepts are now part of university curricula on digital architecture and design ethics, signaling a shift in how spatial environments are conceived through virtual and emotional parameters.
The work of other contemporary artists like Anna Ridler, Robbie Barrat, Claire Silver and Mario Klingemann is heavily informed, produced and curated through machine learning. Yet, the training parameters, aesthetics and curatorial choices remain theirs, underlining the human role in shaping the machine’s creative direction.
Redefining authorship
These examples highlight a deeper shift happening beneath the surface. With A.I. capable of painting like a master, designing like an architect or composing music indistinguishable from human-made works, we must reconsider what makes art “authentic.” Is it the human touch and guidance, the originality of the idea, or perhaps the collaboration itself that defines authenticity? Artists have been using A.I. to create art for half a century already. Harold Cohen began developing his AARON program, a rule-based system that created art autonomously, in 1972, and Lillian Schwartz began creating computer-based visuals in the 1960s—both pioneers in machine-assisted creativity.
Today, artists such as Sougwen Chung, who paints alongside robotic systems, and Sofia Crespo, whose neural network-generated imagery explores biodiversity and synthetic life, are expanding the aesthetic and conceptual range of A.I. as a medium. Their work shows that authenticity is no longer tied solely to human intention, but to the evolving dialogue between human and machine. In this expanded field, authorship becomes distributed, and creativity emerges as a hybrid force—shaped as much by data, training sets and code as by intuition, emotion and experience. What we are witnessing is not just a new tool, but a redefinition of the creative act itself.
Legal, ethical and market implications
A pivotal moment ignited in 2018 when Christie’s auctioned the A.I.-generated Portrait of Edmond de Belamy for $432,500 (a massive increase from its $10,000 presale estimate). Although the price was disproportionate to any A.I. artwork sold before, it started a global dispute around the value and authorship of A.I.-generated works. The sale marked a turning point—it was no longer just artists and coders experimenting in niche circles. Suddenly, the market had validated A.I. as a potentially lucrative creative force, signaling that audiences and institutions might be ready to accept machines as legitimate co-authors. It also forced the art world to confront pressing legal and ethical questions around copyright, authorship and value. From that sale, A.I. art was no longer theoretical—it had arrived on the global stage with market legitimacy.
Of course, there are concerns. Many critics argue that A.I. will undermine human labor and lead to creativity devaluation, which in turn will create repetitive or formulaic art. But if we look harder, history dictates otherwise. Photography did not destroy painting; it pushed artists toward modernism. The same is true with the print and text-based outputs. In the same manner, A.I. comes with new challenges, though we argue that this creates more possibilities than limitations.
Importantly, A.I. creativity sources from, and blends together, different spheres such as art, science, philosophy and technology, and facilitates artists, technologists and scientists to share knowledge in previously unimaginable ways. This merging of disciplines is key, as it opens up more creative thinking, blurring the boundaries of established guidelines and research principles and cultivating deeper cultural dialogue across space and time. In short, it breaks down traditional barriers and silos of knowledge and encourages deeper cross-culture and cross-sector conversations.
Among the most interesting features of machine learning (of which A.I. is a part, but not the whole) is its capacity to create beyond the bounds of human cognition. While human creativity is shaped and limited by the culture, education and personal perspective. A.I. is not. That said, currently available large language models (LLMs) often reflect the very human biases and prejudices embedded into their training data. As more research emerges, these limitations are drawing increased scrutiny. Intellectual property concerns also loom large. The commercial art domain in particular raises important issues of attribution and ownership, as the participation of galleries, collectors and institutions dealing with A.I. art creates an urgent need for well-defined policies on rights and royalties. These vital problems must be resolved to enable the responsible adoption of A.I. into our cultural economy.
Cultural institutions respond to the challenge
Some institutions are beginning to respond. For example, the Serpentine Galleries in London launched their Future Art Ecosystems initiative, which proposes new infrastructures for art in the age of emerging technologies, including guidelines for commissioning, ownership and co-authorship in A.I.-driven works. Similarly, curators at ZKM Center for Art and Media Karlsruhe have incorporated ethical frameworks into their A.I. exhibitions, actively involving interdisciplinary advisory boards to assess both aesthetic value and social implications. While imperfect, these initiatives mark a growing awareness that curating A.I. art is not just about showcasing innovation, but shaping its cultural integration.
It places an obligation on artists, curators and cultural leaders to thoughtfully critique and direct how these technologies are used towards creative practices, well before legislators and regulatory bodies establish a framework. Art is a famously slippery beast to tackle with protocols and laws.
Curating the future: A.I. as a partner in cultural innovation
Despite the above pros and cons still being evaluated, as curators, we maintain that this cross-fertilization of the knowledge sphere can provide fresh sources of inspiration and uncover artistic possibilities that artists haven’t yet imagined. A.I. brings a distinct creativity to the table. Its outputs are continually reshaping how we understand and experience art and culture. These collaborations, in their purest form, aspire toward a more collective vision of creation.
Our approach then is geared primarily towards viewing A.I. not as a rival, but as a companion in the artistic and cultural endeavor, working not against but alongside humans. For fellow curators entering this space, we suggest beginning with openness: embrace experimentation, engage with technologists early and prioritize transparency around datasets and authorship. Co-creating with A.I. means asking different questions—about process, control and intention—and being willing to let the unexpected emerge. It’s less about curating objects and more about curating systems of interaction, where human values and machine logic can meet. History will tell if we got this right.
