The impact of artistic practice on AI development
As both artificial intelligence (AI) and creativity are being foregrounded in UK policy agendas, this article identifies a striking underrepresentation of artists and artistic practice in cultural policy discussions around creative innovation. Despite increasing academic literature, arts-led research, and case studies evidencing a close and dialogic relationship between art and AI, artistic practice remains largely overlooked in policy frameworks addressing the intersection of the cultural sector, creative industries, and digital sector.
To illustrate this, we first call attention to the significant impact artistic practice has on the development of AI, in contrast to the more common discourse of AI’s impact on the arts. Across disciplines, artists have driven the technical advancement of AI tools and techniques, contributed to cross-disciplinary creative technology ecosystems, and informed critical discourse around the ethical implications of the technology.
Artists bridge disciplinary silos, creating spillover effects that drive innovation not just within the arts, but across adjacent industries. Sculptors prototype forms using modelling software, visual artists modify code to remodel AI-generated images, and multi-disciplinary artists produce generative immersive realities (National Endowment for the Arts, 2021; Ploin et al., 2022). British choreographer Wayne McGregor’s “Living Archive,” for example, in collaboration with Google Arts & Culture, explores dance as code, and vice versa, feeding thousands of archival videos into an artificially intelligent tool for audiences to construct choreographic experiments (McGregor, 2019).
Artists also afford a critical space for interrogating the impact of advanced technologies in society. Art often results in AI being deployed in new and unexpected ways, enabling nuanced and evolving engagement with ethical questions (Stark and Crawford, 2019), and opening up space for user intervention, public scrutiny, and policy debate (Hemment et al., 2023). Artist Agrawal (2016) exposes the overrepresentation of Western art in datasets by drawing on Indian art forms to reimagine generative AI with a more diverse, locally-grounded aesthetic. Addressing a different kind of underrepresentation, Ginsberg (2024) questions human-centric design biases by collaborating with horticulturalists, pollinator experts, and a computer scientist to develop “Pollinator Pathmaker,” a living artwork algorithmically optimised for pollinator biodiversity.
The underrepresentation of artists in cultural policy
This artistic practice with AI is taking place against the backdrop of a welcome and important recognition in policy of the value that the creative industries bring as drivers of technological development. The UK aims to position itself over the next 10 years as the best place to live and work with AI, with the creative industries identified as a core sector to achieve this goal (UK Office for Artificial Intelligence, 2021; Department for Digital, Culture, Media and Sport, 2023).
However, across the policy reviewed for this article, there is a clear priority placed on industry as a driver of creative innovation, with a notable under-recognition of artists and non-commercial artistic practice as contributors to the same or similar innovation agendas. The underrepresentation of artists reveals a misalignment between current cultural policy promoting creative innovation and the creative development of AI in practice.
The work of artists discussed in this article, and others, drive the creative innovation of AI models, tools, and datasets. Furthermore, the artworks are realised at the intersection of industries and disciplines that span the cultural sector, creative industries, and digital sector. These characteristics, also underscored in literature (Galleries, 2020; National Endowment for the Arts, 2021; Ploin et al., 2022), support the argument that artistic practice with AI has spill-over effects on innovation in art-adjacent industries (Serpentine Galleries, 2020). However, cultural policy appears to overlook this contribution of artists to the same ecosystem of creative innovation it seeks to promote, despite practice and literature evidencing the contrary.
This underrepresentation suggests a risk of cultural policy not only leaving behind key contributors to technological development, but also not fully capitalizing on the value that artistic practice brings to the development of AI through nuanced critical engagement with the technology. Mitigating this risk requires a clear recognition in cultural policy of the value that art and artists bring to the development of AI, allowing artistic practice engaging with AI to be more formally embedded in policy and infrastructure designed to support innovation.
Recalibrating policy language to better represent artists
Standardised definitions provide the foundations for much policy, therefore, this recognition needs to start at a foundational level; with a reconsideration of the language being used to discuss the intersection of the cultural sector, creative industries, and advanced technologies.
Primarily, accurately capturing the nature of artistic practice with AI requires semantics that move away from disciplinary silos to embrace ecosystems, and foreground process over output. The Standard Industrial Classifications (DCMS and DSIT, 2023) illustrate the semantic separation of art and technology in the language on which much cultural policy is based. Whilst there is a recognition of technology as a sub-sector of the “Creative Industries,” there is no link between this sub-sector and the arts. Further conveying this separation, the definition of the “Digital Sector” recognises the sub-sector “Film, TV, Video, Radio and Music,” but does not recognise “Visual arts.”
As the literature, arts-led research, and case studies evidence, artists have a significant influence on the technical development of AI, and social engagement with the technology. However, taking into account this reality, the SIC context of “Artistic creation” as disconnected from the “Digital Sector” and the lack of overlap between the “Cultural Sector” and “Digital Sector” around “Visual arts” does not reflect this. As a result, artists and artistic practice do not have an accurate or even articulated place within the SIC codes on which much cultural policy regarding innovation is based. Semantically, this reveals a foundational under-representation of artists and artistic practice within the policy discussing or designed to support creative innovation with advanced technologies.
Furthermore, “Artistic creation” itself suggests a misplaced emphasis on output over process. Across literature there is an emphasis that “creation” at the intersection of arts and AI is based on deep inter-disciplinary interrogation, technical trial and error, and long-term collaboration across individuals and sectors. The act of making is therefore as much an act of creation as an embodied process spanning many years, and often full careers of navigating diverse fields and expertise (National Endowment for the Arts, 2021). As such, “Artistic creation” reductively connotes the “creation” of an end product, whilst “practice” more expansively suggests the process required for creation, and the individual artists’ practice honed over the course of a career or lifetime.
This misalignment is likely not the intention of policymakers, and policies such as the much-needed “Creative Industries Sector Vision” (Department for Digital, Culture, Media and Sport, 2023). However, without specifically naming artists as important drivers of technological development in policy documents, the language on which such documents are based subtly yet formally excludes artists. As policy is, by design, authored to effect actionable change, inevitably this will result not only in an exclusion of artists from policy language and literature alone, but also from valuable opportunities and infrastructural support.
Conclusion: Towards an AI-inclusive cultural policy
Cultural policy needs to continue the welcome and necessary investment in the creative industries. However, alongside this, we invite an interrogation of the current semantics being used to articulate the arts, cultural, and creative ecosystems in the context of AI, to align between arts and technological development more closely.
Being able to confidently articulate artistic practice as innovating AI tools and techniques, contributing to cross-disciplinary creative technology ecosystems, and informing critical technology discourse is a first step in developing cultural policy that recognises, prioritises, and invests in artists, and the cultural institutions that support them, as the agents of technological development they are.
