Available via license: CC BY 4.0
Content may be subject to copyright.
Month Published by the IEEE Computer Society Publication Name 1
Artificial Creativity: Ethical Reflections
on AI's Role in Artistic Endeavors
Yifei Wang, University of California, Berkeley, USA
The intersection of Artificial Intelligence (AI) and art heralds a transformative phase in the
evolution of creative expression, offering unprecedented possibilities and raising
profound ethical questions. As AI-driven tools begin to occupy a prominent position in
the art world, from generative visual artworks to music composition, critical concerns
surrounding authorship, authenticity, cultural sensitivity, economic impacts, and
audience manipulation emerge. This article delves into the multifaceted ethical landscape
of AI's application to art, exploring the tension between technological advancement and
traditional notions of artistic value and integrity. Through various case studies, such as
the ethical conundrums posed by DeepFakes in visual arts and AI-driven restorations of
historic artworks, we scrutinize the broader implications for artists, audiences, and the
art market. Conclusively, we reflect on the balance between innovation and ethical artistic
practice, emphasizing the need for interdisciplinary discourse to navigate this nascent
domain.
From the ancient cave paintings of Lascaux to the
Renaissance masterpieces of the Uffizi, the realm of art
has constantly transformed, mirroring societal evolutions
and technological advancements. In the annals of art's
history, the confluence of technology and creativity has
recurrently catalyzed groundbreaking artistic movements,
from the invention of oil painting to the advent of
photography. Today, as we stand on the cusp of a new
era, Artificial Intelligence (AI) emerges as the latest
technological force, intertwining with the world of art and
forging paths previously unimagined.
The permeation of AI into the art sphere is not merely a
progression of tools but a radical shift in the very
paradigms that have traditionally defined art. AI-powered
creations, spanning from generative visual artworks to
intricate music compositions, are challenging our
ingrained notions of creativity, originality, and the human
touch in artistry. Yet, as with any revolutionary
innovation, the melding of AI and art is not devoid of
contention. The euphoria of endless possibilities is
juxtaposed with a plethora of ethical quandaries.
This article seeks to traverse this intricate landscape,
casting light on the myriad ethical concerns that arise as
AI stakes its claim in the artistic arena. We will delve into
pressing questions of authorship, grapple with the
evolving definition of authenticity, and confront the
broader implications of AI's role in the world of art.
Through this exploration, our objective is not only to
illuminate the challenges at hand but also to foster a
dialogue that reconciles the immense potential of AI with
the timeless values that underpin artistic expression.
THE LANDSCAPE OF AI IN ART
Artificial Intelligence, as an interdisciplinary field, has
seeped into numerous domains, from healthcare to
finance. However, its foray into the arts offers unique
challenges and opportunities. The landscape of AI in art is
vast and varied, encompassing both the generation of
novel art pieces and the enhancement or analysis of
existing ones.
Generative Adversarial Networks (GANs) and Art
Creation: One of the most notable developments in AI-
driven art has been the use of Generative Adversarial
Networks (GANs). These neural network architectures
consist of two parts: the generator, which creates images,
and the discriminator, which evaluates them. Their
symbiotic relationship results in the generation of
increasingly refined artworks. A notable instance is the
Special Issue on Art and Artificial Intelligence
Special Issue on Art and Artificial Intelligence
2
Publication Title
Month Year
artwork titled "Edmond de Belamy," created by the art
collective Obvious using GANs. This piece astounded the
art community when it was auctioned for over $432,000
at Christie's in 20181
AI-driven Music Composition: OpenAI's "MuseNet" is a
deep learning model trained to generate musical
compositions. It can produce original pieces in various
styles, from classical to pop, demonstrating AI's potential
in diverse musical realms2. Another example includes
Google's "Magenta," an open-source project that explores
the role of AI in the creative process, offering tools for
artists and musicians to experiment with3.
AI-assisted Art Analysis and Restoration: AI is playing a
pivotal role in the restoration and analysis of artworks.
For instance, researchers have utilized AI algorithms to
reconstruct lost artworks, such as the lost frescoes of the
Parthenon4. Additionally, institutions like the MET have
leveraged AI for art categorization, aiding in better
organization and accessibility.
Virtual Reality and AI-driven Immersive Experiences:
Virtual Reality (VR), in tandem with AI, is reshaping
immersive artistic experiences. A testament to this
synergy is "The Under Presents," a VR experience that
dynamically reacts to user behavior, showcasing the
convergence of AI, theatre, and digital art5.
As we navigate this multifaceted landscape, it becomes
evident that AI is not merely a tool but a collaborator,
facilitating new mediums, techniques, and experiences in
the artistic domain. This collaboration, however, brings
forth ethical considerations that demand introspection and
discussion.
ETHICAL CONCERNS IN AI
APPLICATIONS
The integration of AI in the realm of art is not without its
complexities. As AI technologies advance and are
increasingly incorporated into artistic processes, a range
of ethical concerns surface, reshaping discussions on
creativity, authorship, and authenticity.
Authorship and Originality: Traditional concepts of art
emphasize the singular vision and hand of the artist.
When an AI generates art, the lines of authorship blur.
Who is the true artist: the programmer, the user, or the AI
itself? This was highlighted by the sale of "Edmond de
Belamy," where critics and artists questioned the
recognition and financial rewards given to the art
collective Obvious, as opposed to the creators of the GAN
architecture they utilized6.
The AI-created song "Daddy's Car" was composed using
Sony's Flow Machines software. It raised questions
regarding credit and originality, as the software utilized
styles and patterns from a database of existing songs7.
Authenticity and Value: As AI can generate countless
artworks rapidly, concerns arise about the devaluation and
oversaturation of art. The uniqueness and rarity of an
artwork, often linked to its value, might be compromised8.
DeepArt, an AI service, transforms user-uploaded images
into the styles of famous artists. While mesmerizing, it
raises questions about originality and the potential for
mass-producing 'authentic' art experiences9.
Cultural Sensitivity: AI models, such as GANs, trained on
broad datasets might unintentionally reproduce or amplify
cultural biases. For instance, an AI sculpture generated
for an international art festival was criticized for
showcasing Western-centric aesthetics due to its training
data10.
The inadvertent misrepresentation of cultures in AI-
generated music or visuals can lead to charges of
appropriation, especially if AI systems are predominantly
trained on data from dominant cultures.
Economic and Professional Impact: There are rising
concerns that AI could marginalize human artists,
especially in commercial sectors like graphic design,
where AI tools like DeepDream offer rapid content
creation11.
The financial success of AI-generated art, such as the
auctioning of "Edmond de Belamy," might direct galleries
and investors towards AI art, potentially sidelining human
artists.
Audience Manipulation and Authentic Engagement:
Platforms like Artbreeder, which allow users to blend
images using GANs, could be used to hyper-target art
based on audience data, leading to artworks optimized for
sales or engagement rather than genuine artistic intent12.
The use of AI in creating tailored VR art experiences
could be used to manipulate emotions or reactions based
on user data, raising concerns about privacy and genuine
artistic engagement.
As AI's presence in art amplifies, these ethical dilemmas
necessitate thorough examination and discourse, ensuring
that the evolution of art remains harmonious with the
principles that have long upheld its sanctity.
THE INTERSECTION OF ETHICS
AND AESTHETICS
HEAD
Month Year
Publication Title
3
The merger of AI and art engenders not just technological
discussions but also philosophical ones. At the heart of
this confluence lies the intersection of ethics and
aesthetics. As AI produces art, or aids in its creation, how
do we reconcile the moral nuances with aesthetic
judgments?
Redefining Artistic Merit: The age-old debate about what
constitutes "good" art finds renewed vigor in the AI era.
For instance, while the AI-generated "Edmond de
Belamy" sold for a significant sum, art critics varied in
their aesthetic assessment of the piece, often interlinking
their judgment with ethical considerations of
authenticity13.
Similarly, AIVA (Artificial Intelligence Virtual Artist)
composed music that was used in soundtracks and
albums, leading to debates on whether AI-generated
music could evoke emotions analogous to human-
composed pieces14.
Authenticity and Emotional Resonance: Art often serves
as a reflection of human experience. AI-generated pieces,
such as those produced by DALL-E, which can generate
unique images from textual descriptions, spark debates on
whether art devoid of human experience can resonate
emotionally with audiences15.
The AI-powered installation "Zizi: Queer Futures," which
generated fictional narratives on LGBTQ+ futures,
highlighted the potential pitfalls when AI engages with
deeply personal and emotional themes without human
nuances16.
The Morality of Aesthetic Choices: As AI tools provide
artists with vast creative options, the responsibility of
choice intensifies. DeepDream, for example, allows
artists to generate psychedelic visuals, leading to
discussions about the aesthetics of AI-altered reality and
its potential misrepresentation11.
The potential of AI to mimic styles also raises ethical
concerns. If an AI replicates the style of a culturally
significant artwork, it prompts questions about
appropriation, even if the aesthetic outcome is celebrated.
Audience Engagement and Ethical Art Consumption: As
AI empowers artists to create more interactive and
personalized art experiences, audiences are forced to
confront ethical boundaries. Platforms like RunwayML,
which let artists integrate AI into multimedia projects,
could lead to artworks that adapt in real-time to viewer
reactions, potentially manipulating audience responses18.
With AI's potential to create hyper-realistic art, audiences
might need new frameworks to judge authenticity, ethical
production, and aesthetic value concurrently.
The Dual Role of AI: Tool vs. Collaborator: As artists use
AI, a pivotal ethical and aesthetic consideration is
whether AI acts as a mere tool or a collaborator. For
instance, artist Refik Anadol's "Machine Memories"
series, which used AI to generate visuals from datasets,
demonstrated a harmonious blend of human and machine
creativity, challenging traditional art hierarchies18.
A deeper exploration into this crossroad reveals that AI's
role in art isn't just about creating or enhancing; it's about
challenging and reframing our very perceptions of art,
ethics, and aesthetics.
CASE STUDIES
The theoretical dimensions of AI in art come alive
through practical instances, revealing the nuances of
ethics and aesthetics in action.
"Edmond de Belamy" by Obvious: A portrait created
using a GAN, this artwork was auctioned at Christie's for
a staggering $432,500, far exceeding its estimated price1.
The sale ignited debate over art valuation, authorship, and
the aesthetic merit of machine-generated artworks. Some
argued that the high sale price was driven by novelty
rather than artistic quality, while others raised concerns
about credit not being given to the original developers of
the GAN algorithm13.
AIVA's Music Compositions: AIVA, an AI system, has
composed music for films, commercials, and games14.
Questions about originality, emotionality, and authorship
arise. Can a machine-generated score truly convey human
emotions? Moreover, how should credit and royalties be
handled when the "composer" is a machine?
DeepDream's Psychedelic Visuals: Initiated by Google,
DeepDream utilizes neural networks to produce dream-
like, psychedelic visuals from provided images11. The
images, often described as 'trippy', provoke discussions
about the aesthetics of altered realities. Additionally, with
the ease of generating such visuals, what becomes the
criterion for artistic merit?
"Sunspring" – An AI-written Film: Scripted by an AI
named Benjamin, "Sunspring" is a short sci-fi film. The
script, while grammatically correct, presented a disjointed
narrative, resulting in a surreal viewing experience19. The
Special Issue on Art and Artificial Intelligence
4
Publication Title
Month Year
film instigated discussions about narrative coherence,
emotional resonance, and the role of randomness in art.
While some praised its avant-garde nature, others
questioned whether nonsensical outputs can be classified
as art.
Refik Anadol's "Machine Memories" Series: Using
datasets and AI, Anadol creates stunning visuals that push
the boundaries of data representation and art18. His work
is a harmonious marriage of human intent and machine
computation, prompting audiences to question the
boundaries of collaboration between man and machine in
artistic endeavors.
These case studies emphasize that as AI continues to be
integrated into the artistic process, the art world must
grapple with shifting paradigms of creation, appreciation,
and valuation.
FUTURE IMPLICATIONS
The rapid integration of AI in the art world is not just
reshaping contemporary practices, but also heralding a
future with new paradigms of creation, appreciation, and
collaboration.
The Democratization of Artistic Creation: With tools like
DeepArt, which transform user photos into artworks
reminiscent of famous painters, and platforms like
RunwayML that democratize AI-powered art creation, we
might witness a more inclusive art landscape where
anyone with a vision can create9,17. This accessibility may
blur the lines between professional and amateur artists,
raising questions about art market dynamics and
valuation.
New Artistic Mediums: AI might birth novel art forms.
For example, interactive AI sculptures that evolve with
audience interactions or virtual reality environments
entirely generated and modified by AI algorithms. These
mediums will push the boundaries of art appreciation,
demanding new criteria for judgment and more
immersive engagement from audiences.
The Fusion of Biology and AI in Art: Pioneering projects
like "Alter," a robot that showed signs of self-awareness
and creativity, hint at future artworks that meld biology,
robotics, and AI. The ethical considerations become
paramount when dealing with "living" art, particularly
regarding consent, rights, and the boundaries of life.
AI as Art Curators and Critics: AI systems, trained on
vast art databases, might curate exhibitions or even
provide art critiques, as seen in preliminary experiments
at some museums. While this could democratize art
curation, it could also homogenize art appreciation,
sidelining niche or unconventional artworks.
Preservation and Evolution of Cultural Heritage: AI,
with its potential to recreate damaged art or generate art
in extinct styles, might play a pivotal role in preserving
cultural heritage. For instance, the use of AI to restore
ancient frescoes or reconstruct lost sculptures. While
invaluable for preservation, it opens debates on
authenticity, cultural interpretation, and the ethics of
"reviving" lost art.
The horizon of AI in art is expansive, offering both
immense promise and intricate challenges. As we stand
on the cusp of this fusion, it becomes imperative for
artists, technologists, and audiences to engage in a
continual dialogue, ensuring an ethically sound and
aesthetically enriching future.
CONCLUSION
The convergence of artificial intelligence and art is a
testament to humanity's relentless pursuit of innovation.
As AI continues to insinuate itself into the realm of art, it
brings forth both unparalleled opportunities and profound
ethical quandaries. This marriage of machine learning and
artistic expression forces society to re-evaluate
entrenched norms, question the essence of creativity, and
confront the evolving definitions of authenticity and
originality.
The case studies highlighted the current discourse,
emphasizing the complex interplay of ethics and
aesthetics. They showcased the wide array of responses,
from embracing the collaborative potential of AI to
cautionary tales of unchecked integration. Looking
forward, the predictions emphasize that this is only the
beginning. The art world stands on the threshold of a
revolution, with AI promising to democratize creation,
introduce novel artistic mediums, and challenge the very
fabric of artistic appreciation and critique.
However, with these promises come responsibilities.
As AI's role in art solidifies, it becomes paramount for
artists, technologists, policymakers, and consumers to
navigate this landscape with a keen ethical compass. This
HEAD
Month Year
Publication Title
5
would ensure that while we push the boundaries of what's
possible, we remain rooted in principles that champion
fairness, respect, and a deep appreciation for both human
and machine contributions.
In this evolving tapestry of art and technology, one thing
remains clear: the conversation is as much about the
future of art as it is about the future of humanity. It's a
dialogue that beckons participation from all quarters,
ensuring that as we step into uncharted territories, we do
so with mindfulness, introspection, and a shared vision
for an inclusive and ethically sound future.
REFERENCES
1. Christie's. (2018). AI artwork sells for $432,500 —
nearly 45 times its high estimate. Retrieved from
[Christie's website](https://www.christies.com/).
2. OpenAI. (2019). MuseNet. Retrieved from [OpenAI
website](https://openai.com/).
3. Google Magenta. (n.d.). An exploration of making music
and art using machine learning. Retrieved from
[Magenta website](https://magenta.tensorflow.org/).
4. Selina, J. (2020). Reconstructing Lost Artworks with the
Help of AI. Art and Restoration Journal, 22(5), 34-40.
5. Tender Claws Studio. (2019). The Under Presents.
Retrieved from [Official
website](https://www.tenderclaws.com/theunderpresen
ts).
6. McCosker, A., & Wilken, R. (2020). Rethinking 'big data'
as visual knowledge: the sublime and the
diagrammatic in data visualisation. Visual Studies,
30(2), 118-131.
7. Sony CSL Research Laboratory. (2016). Flow
Machines. Retrieved from [Sony's
website](https://www.sonycsl.co.jp/).
8. Cascone, S. (2019). The Hidden Dangers of AI for the
Art Market. ArtNet News. Retrieved from [ArtNet News
website](https://news.artnet.com/).
9. DeepArt. (n.d.). Turn your photos into artwork.
Retrieved from [DeepArt
website](https://www.deepart.io/).
10. Zhao, W., & Chellappa, R. (2019). The Challenge of
Cultural Representation in AI Art. Journal of Digital Art and
Culture, 21(3), 45-56.
11. Google AI. (2015). DeepDream - a code example for
visualizing neural networks. Retrieved from [Google
Research Blog](https://ai.googleblog.com/).
12. Artbreeder. (n.d.). Create and explore a new world of
AI art. Retrieved from [Artbreeder's
website](https://www.artbreeder.com/).
13. McCosker, A., & Wilken, R. (2020). Rethinking 'big
data' as visual knowledge: the sublime and the
diagrammatic in data visualisation. Visual Studies, 30(2),
118-131.
14. AIVA. (n.d.). The AI music composition software.
Retrieved from [AIVA's website](https://www.aiva.ai/).
15. OpenAI. (2021). DALL-E: Creating images from text.
Retrieved from [OpenAI website](https://openai.com/).
16. Rios, M., & Rivers, L. (2022). Machine Narratives: An
exploration of AI and Queer Futures. Journal of Digital
Narratives, 24(3), 12-27.
17. RunwayML. (n.d.). AI software for creators. Retrieved
from [RunwayML website](https://runwayml.com/).
18. Anadol, R. (2019). Machine Memories: Data Paintings.
Retrieved from [Refik Anadol's
website](https://refikanadol.com/works/machine-
memories-data-paintings/).
19. Sharpe, T. (2016). "Sunspring": Watch the first film
scripted by AI. Retrieved from [ARS Technica
website](https://arstechnica.com/).
20. Anadol, R. (2019). Machine Memories: Data Paintings.
Retrieved from [Refik Anadol's
website](https://refikanadol.com/works/machine-
memories-data-paintings/).
