Scholarly record
ARTIFICIAL INTELLIGENCE AS ARTIST AND HISTORIAN: INTERPRETING AND INVENTING STYLES IN THE AGE OF MACHINE CREATIVITY
Abstract
Generative Artificial Intelligence (Gen-AI) is redefining both the creation of art and the interpretation of cultural heritage. In this paper, we conceptualise AI in a dual role: (1) Art Historian, using AI to reconstruct, reimagine, and annotate historical artworks; and (2) Artist, using AI to produce novel machine-native aesthetics that challenge traditional notions of style and authorship. We review recent advances in diffusion models, neural style transfer, and vision–language systems that enable these capabilities. Building on this review, we develop a structured framework for “AI-as-Historian”, encompassing digital reconstruction of damaged or lost art, creative reimaginings of artworks in new styles, and AI-driven annotations or guides for cultural collections, and propose a taxonomy of emerging “latent-space aesthetics”. This taxonomy describes visual tendencies of AI- generated art, including hyperrealistic latent realism, surreal morphing of forms, prompt- driven conceptual art, and model-specific noise signatures. We illustrate these dual roles with conceptual case studies and visual examples, demonstrating how an AI might digitally restore a damaged painting or re-envision a masterpiece in another style, alongside the entirely new artistic genres born from generative models. We further discuss the ethical and curatorial implications of AI-mediated art interpretation, such as questions of authenticity, bias in training data, and interpretive authority, and outline future research directions on AI explainability, cross-cultural representation, and audience engagement. Our conclusions highlight that integrating AI into art history and practice offers exciting opportunities to deepen public engagement with art, provided it is done responsibly and with critical awareness of AI’s limitations.
Publication Impact Profile
Publication details
References15
Baum J., Villasenor J., Rendering misrepresentation: Diversity failures in AI image generation, Brookings Institution, United States, 2024. URL: https://www.brookings.edu/articles/rendering-misrepresentation-diversity-failures-in-ai- image-generation
Neural Style Transfer, Wikipedia, United States, (last accessed on October 2025),URL: https://en.wikipedia.org/wiki/Neural_style_transfer
How can AI contribute to art historical analysis and research, EdenAI, (2025), URL: https://www.edenai.co/post/how-can-ai-contribute-to-art-historical-analysis-and- research
Fenstermaker W., How Artificial Intelligence Sees Art History, The Metropolitan Museum of Art Perspectives, United States, February 4, 2019. URL: https://www.metmuseum.org/perspectives/artificial-intelligence-machine-learning-art- authorship
Obvious and the interface between art and artificial intelligence, Christie’s, United States, 12 December 2018. URL: https://www.christies.com/en/stories/a-collaboration- between-two-artists-one-human-one-a-machine-0cd01f4e232f4279a525a446d60d4cd1
Chu J., Have a damaged painting? Restore it in just hours with an AI-generated mask, MIT News, United States, June 11, 2025. URL: https://news.mit.edu/2025/restoring- damaged-paintings-using-ai-generated-mask-0611
How Can AI Be Used in Heritage Conservation Projects, Orbit-O-R, June 10 2025. URL: https://www.orbit-o-r.com/post/how-can-ai-be-used-in-heritage-conservation- projects
Gonsalves R.A., Using GPT-4 with Vision as an Art Critic, TDS Archive / Medium, United States, Nov 2 2023. URL: https://medium.com/data-science/using-gpt-4-with- vision-as-an-art-critic-ec91080ba334
Hyperrealism, Wikipedia, United States, (last accessed October 2025), URL: https://en.wikipedia.org/wiki/Hyperrealism_(visual_arts)
Synthetic Reality Reimagined: From Hyperreal to SuperReality, Psychology Today, United States, October 2023. URL: https://www.psychologytoday.com/us/blog/the- digital-self/202310/synthetic-reality-reimagined-from-hyperreal-to-superreality
Bignotti F., Potential bias in AI: cultural representation and the marginalization of African art, University of Verona, Italy, 2025. DOI / additional identifier if available: https://aiucd2025.dlls.univr.it/assets/pdf/papers/7.pdf
Millet K., Buehler F., Du G., Kokkoris M. D., Defending humankind: Anthropocentric bias in the appreciation of AI art, Computers in Human Behavior, United States, vol. 143, 2023, Article 107707. DOI: 10.1016/j.chb.2023.107707
Yellowbrick, Strategies for Bias Identification and Mitigation in AI Art, Yellowbrick blog, United States, July 25, 2024. URL: https://www.yellowbrick.co/blog/animation/strategies-for-bias-identification-and- mitigation-in-ai-art
Islam A., How Do DALLВ·E 2, Stable Diffusion, and Midjourney Work?, MarkTechPost, United States, November 14, 2022. URL: https://www.marktechpost.com/2022/11/14/how-do-dall%C2%B7e-2-stable-diffusion- and-midjourney-work/
Google DeepMind, Gemini: A Family of Highly Capable Multimodal Models, United States, 2023. arXiv preprint arXiv:2312.11805. DOI: 10.48550/arXiv.2312.11805
View or Download full articleAccess options
SWS access login
Login as SWS Scientific CommitteeLogin as SWS Scientific PartnerLogin as SWS AuthorAuthors and approved SWS contributors will read and export their own linked papers after identity matching by SWS profile, email and SGEM GlobalID.
For librarian assistance: [email protected]
Purchase Instant Access
- Article can be downloaded after successful payment.
- Article may be used according to SWS library access terms.
- Article cannot be redistributed.