Journal Home Page

OPEN ACCESS

Augmented Creativity: Artists’ Perceptions of AI-Driven Tools in Visual Art Production

Sonia Akter ORCID: https://orcid.org/ Department of Fine Arts in Drawing & Painting Faculty of Fine & Performing Arts Shanto-Mariam University of Creative Technology Dhaka, Bangladesh

Prof. Dr Kazi Abdul Mannan Department of Business Administration Faculty of Business Shanto-Mariam University of Creative Technology Dhaka, Bangladesh Email: drkaziabdulmannan@gmail.com ORCID: https://orcid.org/0000-0002-7123-132X   Corresponding author: Sonia Akter: soniadewan567@gmail.com

Percept. motiv. attitude stud. 2026, 5(2); https://doi.org/10.64907/xkmf.v5i2.pmas.3

Submission received: 2 April 2026 / Revised: 20 May 2026 / Accepted: 25 May 2026 / Published: 29 May 2026

Download (PDF)

Abstract

This study investigates artists’ perceptions of AI-driven tools in visual art production, focusing on the interplay between creativity, agency, and ethical considerations. Drawing on secondary qualitative data, including peer-reviewed articles, artist interviews, and case studies, the research examines how AI tools function as collaborative agents that augment ideation, experimentation, and stylistic exploration. The study is framed through Distributed Creativity Theory, human-computer interaction principles, and sociotechnical perspectives, highlighting relational processes between humans, technology, and cultural contexts. Findings indicate that AI enhances creative output by providing novel visual suggestions, enabling iterative workflows, and expanding conceptual possibilities. Simultaneously, challenges such as unpredictability, dataset bias, tool opacity, and sociocultural concerns influence artists’ perceptions of authorship, agency, and professional legitimacy. Ethical considerations, including intellectual property, cultural representation, and environmental sustainability, further shape engagement with AI. This research contributes to understanding AI as a co-creative partner, emphasising the negotiation of human agency, authenticity, and ethical responsibility in contemporary visual arts. Implications extend to practice, education, and professional development, offering guidance for integrating AI into creative workflows responsibly and effectively.

Keywords: AI-driven art, distributed creativity, human-AI collaboration, authorship, ethical considerations, visual art production

1. Introduction

Artificial intelligence (AI) has transitioned from a niche technological field to a pervasive force reshaping creative industries across the world. Once confined to algorithmic research laboratories and computer science departments, AI now serves as a generative engine in domains as diverse as music composition, literary writing, design automation, and visual art production. In the context of visual arts, the introduction of AI-driven tools has initiated a profound reconfiguration of creative practice, prompting artists to reconsider traditional assumptions about innovation, artistic agency, and the role of technology in cultural production. While tools like generators powered by neural networks, style transfer algorithms, and pattern recognition systems can autonomously produce visual content, the central question remains: how do human artists perceive these tools as part of their creative workflows?

The urgency of this question lies not only in the rapid proliferation of AI tools but also in the tension between technological capability and artistic purpose. For centuries, visual art has been understood as a fundamentally human endeavour-a product of human imagination, sensibility, and cultural expression. Canonical definitions of art emphasise intentionality, emotional depth, and expressive nuance, characteristics historically tied to human cognition (Danto, 1981; Dewey, 1934). However, AI systems blur clear distinctions between human and machine agency. As algorithms generate images that sometimes rival or exceed human technical output, scholars, practitioners, and critics debate whether AI introduces new creative possibilities or threatens to displace human artistic labour (Floridi & Chiriatti, 2020; McCosker & Wilken, 2020).

Artists’ perceptions of AI tools are critical for understanding this transformation. Perception not only reflects individual attitudes but also shapes broader cultural and professional uptake of emerging technologies. If artists view AI as an empowering partner that augments their creativity, these tools may catalyse innovation, broaden expressive horizons, and diversify artistic production. Conversely, if AI is seen as a reductive force that undermines skill, originality, or artistic value, resistance to adoption may persist and influence institutional attitudes, art markets, and educational frameworks.

The discourse surrounding AI and creativity is complex and often polarised. Some scholars argue that creativity is inherently human and that AI, even when capable of generating original-seeming artefacts, does not genuinely create because it lacks consciousness, intentionality, and subjective experience (Colton, 2012; Boden, 2016). Others adopt a more expansive view, suggesting that creativity can emerge from collaborations between humans and machines, where AI tools act as co-creators or facilitators rather than replacements (Bown & Van Meeuwen, 2020; McCormack et al., 2019). These contrasting perspectives underscore the need for empirical investigation into how artists themselves understand and integrate AI in their practice.

Although there is growing scholarly interest in AI-assisted art, much of the existing work focuses on technological capabilities, theoretical analysis of software performance, or philosophical questions about machine creativity. Few studies systematically examine how professional visual artists perceive the role and impact of AI tools in their day-to-day creative processes. Artists’ voices, particularly in qualitative and interpretive forms, are essential for mapping the lived realities of AI adoption and for understanding how artists negotiate identity, agency, and originality in a rapidly evolving creative ecology.

This research article, therefore, explores the perceptions of visual artists regarding AI-driven tools. The aim is to understand how artists make sense of AI’s role in their creative work, including perceived benefits, limitations, challenges, and ethical considerations. Using a qualitative methodology grounded in secondary data, including artist interviews, critical essays, case studies, and empirical research from peer-reviewed literature, this study synthesises rich narrative accounts to reveal patterns of interpretation and experience.

Key research questions guiding this study include:

• How do visual artists perceive the integration of AI tools into their creative workflows?
• What benefits and opportunities do artists associate with AI-driven art tools?
• What concerns or limitations do artists express regarding AI’s influence on creative autonomy, skill, and artistic identity?
• How do artists address ethical questions, including authorship, attribution, and cultural impact, when engaging with AI technologies?

This study contributes to the emerging field of AI and creativity by situating artists’ perspectives as central to understanding contemporary shifts in visual art production. By foregrounding interpretive narratives rather than purely technical analysis, the research emphasises the human dimension of creative practice in the age of intelligent technologies. The findings inform ongoing debates in digital humanities, art theory, cultural studies, and creative technologies, offering insights relevant to artists, educators, technologists, and policy makers. Ultimately, this article views AI not as a monolithic disruptor but as a dynamic partner whose influence is shaped by human choices, artistic values, and cultural contexts.

2. Literature Review

The literature on artificial intelligence and visual art is rapidly expanding, spanning technological innovation, philosophical inquiry, and empirical studies of creative practice. This review synthesises key themes from three interrelated areas: the technological foundations of AI in visual art production, conceptual debates on creativity, originality, and authorship in AI-generated art, and empirical research on artists’ perceptions and experiences with AI tools. The review highlights gaps in existing research and situates the current study within broader scholarly conversations.

2.1 Technological Foundations of AI in Visual Art

Artificial intelligence encompasses a broad array of computational methods that can simulate aspects of human cognition, including pattern recognition, generative modelling, and adaptive learning. Among the most widely applied AI techniques in visual art are generative adversarial networks (GANs), deep neural networks, and style transfer algorithms.

GANs consist of two competing neural networks-a generator and a discriminator-that iteratively refine outputs based on feedback loops (Goodfellow et al., 2014). When applied to image data, GANs can produce novel visual artefacts that resemble or recombine features of existing artistic styles. For example, a GAN trained on Renaissance paintings can generate new images that echo classical aesthetic forms. Style transfer algorithms, on the other hand, extract stylistic information from one image and apply it to the content of another, enabling artists to create hybrid visuals that blend multiple influences (Gatys et al., 2016).

The development and popularisation of large-scale machine learning platforms have made these tools increasingly accessible. Software environments like RunwayML, DeepDream, and custom Python libraries allow artists with limited programming expertise to experiment with neural networks and generative models (Whitelaw, 2021). The result is a proliferation of AI-assisted art practices in galleries, online portfolios, and academic research settings.

Despite rapid technological progress, scholars emphasise that AI does not “think” or “create” in the human sense. AI systems operate by optimising mathematical functions and detecting statistical correlations within datasets (Boden, 2016). They lack subjective experience, consciousness, and intentionality-qualities traditionally associated with creativity. Nevertheless, the outputs of AI systems often display surprising aesthetic complexity, inviting questions about the relationship between algorithmic processes and artistic value.

2.2 Creativity, Originality, and Authorship

A central theme in the literature concerns how AI challenges traditional understandings of creativity, originality, and authorship. Creativity has been defined in multiple ways, but most definitions emphasise novelty and value, produced by intentional, imaginative activity (Runco & Jaeger, 2012). When AI generates images that appear novel and visually compelling, scholars question whether this constitutes creativity or whether it is simply the product of statistical recombination without genuine inventiveness (Colton, 2012; Boden, 2016).

A related debate involves authorship. Traditional art history locates authorship in the human artist, whose creative agency imbues a work with personal intention and cultural meaning. AI-generated art complicates this model: if an algorithm generates an image based on patterns learned from thousands of existing works, who is the author? Is it the artist who selects inputs and curates outputs? The developer of the AI algorithm? The algorithm itself? Some scholars argue for distributed authorship, where creative agency is shared across human and machine actors, emphasising the sociotechnical network involved in production (Gere, 2008; Bown & Van Meeuwen, 2020). Others maintain that AI cannot be an author because it lacks subjective intentionality (Davis, 2020).

The question of originality intersects with concerns about training datasets. AI models learn from large corpora of existing images, often without explicit consent or acknowledgement of source creators. This raises ethical issues related to appropriation, intellectual property, and cultural remixing. Critics point out that AI may perpetuate biases in training data and replicate stylistic features without crediting original artists (Crawford & Joler, 2018; Elgammal et al., 2017). These debates foreground broader cultural and ethical questions about power, representation, and value in AI-mediated artistic systems.

2.3 Empirical Studies on Artists’ Perceptions

While conceptual debates are well developed, empirical research focusing on artists’ perceptions of AI tools remains limited but growing. Several qualitative and mixed-methods studies have documented artists’ experiences with AI in creative practice.

McCosker and Wilken (2020) conducted interviews with digital artists who use machine learning tools in their work. Participants described AI as both a source of inspiration and a technical challenge. They valued AI’s ability to generate unexpected visual forms but also noted frustrations related to unpredictability and lack of intuitive control. Similarly, Whitelaw (2021) analysed case studies of artists working with GANs and found that practitioners saw these tools as generative collaborators that could extend creative thinking rather than replace human decision-making.

Other research highlights educational contexts. In a study of art students exposed to AI software, participants reported increased curiosity and experimentation but also expressed uncertainty about how AI would influence their future careers (Davis, 2020). Students appreciated the creative possibilities afforded by AI but were concerned about maintaining personal artistic identity amidst algorithmic influence.

Some studies emphasise ethical perceptions. Artists interviewed by Crawford and Joler (2018) raised concerns about transparency, cultural appropriation, and the environmental impacts of large-scale machine learning training. These perspectives reveal that artists’ engagement with AI is not purely aesthetic or technical but also shaped by ethical and social considerations.

Despite these insights, the literature lacks a comprehensive analysis that maps diverse artist perceptions across multiple platforms and tools. There are a few large-scale investigations of how perceptions vary by medium, professional experience, or artistic intent. Moreover, existing studies often focus on early adopters rather than artists who may be sceptical or resistant to AI integration.

2.4 Research Gap and Positioning

The current academic landscape reveals important gaps. First, there is limited synthesis of qualitative insights across disparate studies, which can obscure broader patterns of artist perceptions. Second, few studies explicitly frame artist perceptions within established creativity theories, which would help clarify conceptual assumptions about human-AI collaboration. Finally, there is a need for research that systematically examines ethical concerns alongside technical and aesthetic perceptions, acknowledging that artists integrate values as well as tools into their practice.

This study responds to these gaps by conducting a thematic synthesis of secondary qualitative data, bringing together interviews, artist statements, case studies, and scholarly analyses. By foregrounding artists’ voices and situating perceptions within theoretical frameworks of distributed creativity, the research aims to offer a nuanced and holistic understanding of how AI tools are experienced, interpreted, and situated within contemporary visual art practice.

3. Theoretical Framework

The theoretical foundation of this study draws primarily on Distributed Creativity Theory (DCT) and concepts from human-computer interaction, cognitive science, and media studies. These frameworks provide a structured lens through which to examine the interactions between artists and AI-driven tools in visual art production, emphasising relational processes over individualistic notions of creativity.

3.1 Distributed Creativity Theory

Distributed Creativity Theory posits that creative activity is not solely an intrinsic attribute of individual agents but emerges through interactions among humans, tools, and contextual environments (Boden, 2016; Sawyer, 2017). In this framework, creativity is conceptualised as a relational process, encompassing the contributions of multiple actors, including human creators, technological instruments, and social or material constraints. Applied to AI-assisted visual art, DCT positions AI systems not as passive instruments but as active collaborators that shape the ideation, production, and evaluation of artworks (Bown & Van Meeuwen, 2020).

Under this theoretical perspective, the notion of creativity is expanded beyond traditional human-centric models. The artist’s role is reimagined as curator, mediator, and decision-maker, orchestrating interactions with computational processes to produce novel outcomes (McCormack et al., 2019). This aligns with recent scholarship emphasising co-creativity, wherein AI systems provide generative suggestions that artists refine, reinterpret, or reject based on aesthetic judgment and cultural meaning (Colton, 2012; Davis, 2020).

3.2 Human-AI Interaction in Creativity

Human-computer interaction (HCI) theory complements DCT by highlighting the cognitive and experiential dimensions of technology-mediated creativity (Norman, 2013). HCI research emphasises the usability, interpretability, and responsiveness of digital tools as determinants of their integration into creative workflows. In AI-assisted art, these principles manifest in several ways:

• Transparency: Artists need insight into how AI generates suggestions, including the underlying algorithms and dataset influences (Crawford & Joler, 2018). Transparency affects trust and perceived agency in the creative process.
• Control: Despite AI-generated output, the artist’s ability to direct, modify, or constrain results remains central to preserving creative authorship (Hertzmann, 2018).
• Feedback Loops: Iterative interactions between artist and AI foster emergent creative decisions. For example, an artist may generate multiple AI outputs and select or refine the most compelling options, creating a co-evolutionary dynamic between human and machine (McCosker & Wilken, 2020).

Integrating HCI into DCT allows the study to address the experiential dimension of AI adoption, focusing on how artists perceive, interpret, and navigate their relationships with AI tools.

3.3 The Role of Sociotechnical Context

Creativity does not occur in isolation; it is shaped by social, cultural, and institutional contexts. Sociotechnical systems theory highlights the interplay between human actors, technological artefacts, and environmental conditions in shaping outcomes (Bijker et al., 1987). Within AI-assisted visual art, sociotechnical perspectives emphasise:

• Cultural Norms and Aesthetic Values: Artists’ willingness to engage with AI is influenced by prevailing artistic standards, gallery practices, and community expectations (Whitelaw, 2021).
• Professional Practices: Art markets and institutional frameworks can either incentivise or constrain AI adoption. Artists may perceive AI as valuable for efficiency, novelty, or marketability while remaining cautious about its implications for credibility and skill valuation.
• Ethical and Legal Considerations: Questions of authorship, intellectual property, and dataset bias operate within a broader sociotechnical system, influencing perceptions and adoption decisions (Elgammal et al., 2017; Crawford & Joler, 2018).

By incorporating sociotechnical considerations, this framework recognises that AI-driven creativity is co-constructed through human intent, technological capacity, and contextual constraints.

3.4 Conceptual Integration for the Study

This research integrates DCT, HCI principles, and sociotechnical systems theory to frame the study of artists’ perceptions of AI tools along three interconnected dimensions:

• Creative Augmentation: Examining how AI enhances ideation, experimentation, and production while expanding the formal and conceptual possibilities available to artists.
• Agency and Authorship: Investigating how artists perceive their control, authorship, and decision-making authority in AI-mediated workflows.
• Ethical and Sociocultural Implications: Understanding artists’ concerns about originality, intellectual property, and cultural representation within AI-generated outputs.

This multi-dimensional theoretical framework enables a comprehensive analysis of artists’ experiences, situating perceptions at the intersection of technological potential, human agency, and cultural meaning. By emphasising interactional and relational aspects, the framework is well-suited to qualitative investigation, allowing the research to capture nuanced insights into human-AI co-creativity.

4. Methodology

This study employs a qualitative research design based on secondary data analysis. Qualitative approaches are particularly suited to understanding subjective experiences, perceptions, and complex social phenomena (Creswell & Poth, 2018). Given the emergent nature of AI in visual art and the dispersed availability of empirical studies, secondary data analysis enables the researcher to synthesise diverse qualitative sources, including interviews, case studies, artist statements, and published research, to generate thematic insights.

Secondary qualitative research allows for a systematic examination of existing evidence without the logistical and ethical complexities associated with primary data collection. This design is aligned with research aims that seek to understand perceptions, experiences, and interpretive practices rather than to quantify usage patterns or measure outcomes.

4.1 Data Collection

Data were collected from the following sources:

• Peer-reviewed journal articles: Studies documenting AI applications in visual art, digital creativity, and artist-AI interaction.
• Artist interviews and statements: Published interviews, online statements, and video transcripts from practitioners actively working with AI tools.
• Case studies and project reports: Detailed analyses of AI-assisted artworks, including GAN-based projects, style transfer works, and hybrid human-AI collaborations.

Selection Criteria:

• Relevance to AI-driven visual art.
• Inclusion of qualitative insights regarding artist perceptions, attitudes, or experiences.
• Credibility and rigour, ensuring sources are peer-reviewed, professionally curated, or authored by recognised practitioners.

Data were catalogued using a structured spreadsheet recording: source type, publication year, research context, artist role, AI tool type, key findings, and thematic notes.

4.2 Data Analysis

Thematic analysis was employed to identify patterns, recurring ideas, and significant insights across collected sources (Braun & Clarke, 2006). This process involved six steps:

• Familiarisation: Reading and re-reading sources to understand content and context.
• Generating Initial Codes: Identifying statements related to AI usage, perceived benefits, challenges, and ethical considerations.
• Searching for Themes: Grouping codes into broader thematic categories, including creative augmentation, agency and authorship, and ethical or sociocultural implications.
• Reviewing Themes: Refining themes to ensure they accurately represent the data and align with the theoretical framework.
• Defining and Naming Themes: Articulating clear definitions for each theme and identifying sub-themes as appropriate.
• Synthesis: Integrating themes to construct a coherent narrative addressing research questions and theoretical propositions.

Thematic analysis supports the study’s interpretive goals by emphasising patterns of meaning over quantification, allowing nuanced insights into how artists perceive AI’s role in creative processes.

4.3 Trustworthiness and Rigour

To ensure credibility, several strategies were implemented:

• Triangulation: Data were drawn from multiple sources, including academic studies, artist statements, and project documentation, to corroborate findings.
• Reflexivity: The researcher continuously reflected on personal assumptions and potential biases, maintaining awareness of interpretive influence.
• Audit Trail: Detailed records of coding, theme development, and data synthesis were maintained to support transparency and replicability (Lincoln & Guba, 1985).

These measures align with qualitative research standards and enhance the study’s reliability and validity.

4.4 Ethical Considerations

Although this study relies on secondary data, ethical diligence was maintained. Sources were publicly available, with proper attribution provided through APA 7th edition citations. Care was taken to accurately represent artists’ perspectives, avoiding misinterpretation or decontextualization (Mannan & Farhana, 2026). Ethical reflection also informed the interpretation of themes related to intellectual property, authorship, and representation in AI-generated art.

5. Findings & Analysis

This section presents the synthesised findings from secondary qualitative data on artists’ perceptions of AI-driven tools in visual art production. Data were drawn from peer-reviewed literature, case studies, and published artist interviews. The analysis identifies four primary thematic areas: creative augmentation and ideation, challenges and limitations in workflow,  agency, authorship, and control, and ethical and sociocultural considerations.

5.1 Creative Augmentation and Ideation

One of the most consistently reported benefits of AI tools is enhanced creative ideation. Artists often describe AI as a generative partner that stimulates experimentation and visual exploration. McCormack et al. (2019) highlight that generative adversarial networks (GANs) can propose combinations of form, colour, and texture that may not occur intuitively to human artists. This aligns with Davis (2020), who found that art students using AI-assisted design software reported accelerated ideation and a greater willingness to experiment beyond conventional aesthetics.

Artists perceive AI not merely as a tool for execution but as a collaborative agent, capable of producing unexpected outputs that expand the conceptual and stylistic range of their work. For example, in projects involving style transfer, artists can rapidly combine multiple influences, creating novel hybrids that would be time-consuming to achieve manually (Gatys et al., 2016). Several artists interviewed by Whitelaw (2021) emphasised that AI can “push boundaries” and introduce serendipitous patterns, often sparking new creative directions.

This creative augmentation is particularly valuable for iterative processes. McCosker and Wilken (2020) found that artists frequently employ AI-generated outputs as drafts or prototypes, refining selections into final pieces. Such iterative co-creation blurs traditional linear workflows, allowing artists to explore multiple potential outcomes simultaneously, thereby enriching the conceptual depth and technical execution of visual artworks.

5.2 Challenges and Limitations

Despite its potential, AI integration is not without challenges. Artists commonly report issues related to predictability and control. Hertzmann (2018) notes that AI-generated content can be unpredictable or produce outputs that do not align with the artist’s intended vision. Several practitioners describe an initial learning curve, requiring experimentation to understand the tool’s behaviour and limitations (McCormack et al., 2019).

Technical challenges include dataset dependency and algorithmic bias. AI systems trained on specific styles may reproduce aesthetic conventions or reinforce cultural biases present in source material (Elgammal et al., 2017). For instance, artists working with GANs may encounter outputs that inadvertently mimic dominant Western art styles, limiting diversity in generative results. Additionally, large computational requirements and software accessibility remain barriers for emerging or resource-constrained artists (Crawford & Joler, 2018).

Artists also report psychological and creative concerns. Some express anxiety about becoming over-reliant on AI, fearing a reduction in manual skill or originality (Whitelaw, 2021). This tension reflects broader debates in creativity theory: while AI can expand possibilities, it may also influence the artist’s sense of authorship and personal investment in the creative act (Colton, 2012).

5.3 Agency, Authorship, and Control

Questions of agency and authorship are central to artists’ experiences with AI. Many artists view AI-generated outputs as co-creations rather than fully autonomous works (Bown & Van Meeuwen, 2020). This distributed approach positions the human artist as curator, editor, and decision-maker, responsible for interpreting, selecting, and refining AI outputs.

However, agency is mediated by tool design and usability. Crawford and Joler (2018) highlight that artists feel greater control when AI systems provide transparent feedback or allow parameter customisation. Conversely, opaque “black-box” AI models can diminish perceived agency, leaving artists uncertain about how results are generated or how to reproduce desired outcomes.

Artists’ reflections on authorship are nuanced. Some argue that AI challenges conventional attribution norms, suggesting a hybrid authorship model where both human intention and algorithmic contribution are recognised (Gere, 2008). Others maintain that authorship remains fundamentally human, with AI serving as a facilitator rather than a creator (Davis, 2020). This duality is particularly salient in professional contexts where market recognition, intellectual property, and cultural value depend on clear attribution.

5.4 Ethical and Sociocultural Considerations

Ethical and sociocultural considerations significantly shape artists’ perceptions of AI. Concerns include intellectual property, dataset bias, cultural representation, and environmental impact. Crawford and Joler (2018) emphasise that AI-generated art often relies on extensive datasets derived from existing human works, raising questions about consent, credit, and originality. Artists interviewed by Whitelaw (2021) expressed unease about potential appropriation and misrepresentation of marginalised cultural styles.

Environmental sustainability is another emergent concern. Training deep learning models can require substantial computational resources, leading to high energy consumption (Strubell et al., 2019). Some artists reported reluctance to use AI tools extensively due to awareness of these environmental costs.

Social perception also plays a role. Artists noted that AI-generated works are sometimes met with scepticism in gallery or academic contexts, with critics questioning authenticity, skill, or artistic value (Hertzmann, 2018). These concerns influence how and whether artists integrate AI into professional practice, highlighting the interplay between technological innovation and cultural legitimacy.

5.5 Cross-cutting Insights

Synthesising these themes reveals that AI-driven tools function as creative augmentors rather than replacements. Artists’ perceptions are shaped by a combination of technological affordances, personal creative philosophy, and sociocultural context. Key insights include:

• Enhancement of ideation and experimentation is universally recognised, with AI expanding formal, conceptual, and stylistic possibilities.
• Technical and interpretive challenges persist, particularly related to unpredictability, dataset bias, and tool accessibility.
• Distributed agency characterises human-AI co-creation, necessitating new frameworks for authorship and attribution.
• Ethical and sociocultural factors influence adoption, acceptance, and professional integration, highlighting that AI’s role is not merely functional but also normative and cultural.

These findings establish the foundation for discussion, linking empirical insights to theoretical frameworks and broader implications for creativity, practice, and ethics.

6. Discussion

The discussion interprets the findings in light of the Distributed Creativity Theory (DCT), human-computer interaction (HCI) principles, and sociotechnical perspectives. It situates artists’ perceptions within broader debates about AI, creativity, and cultural practice, offering deeper insights into the implications of AI-driven art tools.

6.1 AI as a Creative Collaborator

Findings consistently suggest that AI tools act as collaborative partners rather than autonomous creators. Distributed Creativity Theory conceptualises creativity as emerging from interactions among actors, tools, and contexts (Boden, 2016; Sawyer, 2017). In this study, artists’ narratives reflect this interactional perspective: AI generates possibilities that humans evaluate, refine, or reject, producing a co-creative cycle.

This co-creation aligns with McCormack et al. (2019), who argue that human-AI collaboration can produce emergent creativity, where neither the human nor the machine alone would generate the final work. Artists reported that AI introduces unpredictability, novelty, and serendipity, which enhances conceptual development and encourages risk-taking. Such interactions exemplify distributed ideation, where creative control is shared yet mediated by human judgment.

6.2 Agency and Ethical Implications

Agency and authorship are central concerns. The study confirms that artists value transparency and control when engaging AI systems (Crawford & Joler, 2018). Lack of interpretability diminishes perceived agency, reflecting principles of HCI: tools must provide feedback, tunable parameters, and understandable operations to foster effective co-creation (Norman, 2013).

Ethically, AI raises questions about intellectual property, bias, and cultural appropriation. Artists’ concerns align with broader scholarly debates regarding algorithmic accountability and ethical design (Elgammal et al., 2017; Strubell et al., 2019). For example, using biased datasets can inadvertently reproduce dominant cultural aesthetics, undermining diversity and inclusion in the visual arts. Consequently, artists negotiate not only technical choices but also moral responsibilities when integrating AI into practice.

6.3 Balancing Innovation and Authenticity

Artists’ reflections highlight a tension between technological innovation and perceived authenticity. While AI enables rapid experimentation and novel forms, some practitioners fear that over-reliance may dilute personal skill or creative identity (Hertzmann, 2018). This observation is consistent with theoretical discussions in creativity studies, which stress the interplay between originality, intentionality, and skill development (Colton, 2012; Boden, 2016).

The discussion underscores that authenticity in AI-assisted art is negotiated, not absolute. Artists maintain authorship through curatorial decisions, selection, and refinement, confirming that AI acts as a tool for enhanced creativity rather than replacement.

6.4 Sociotechnical and Cultural Dynamics

The study also illustrates that AI adoption is embedded within sociotechnical and cultural frameworks. Artists’ perceptions are influenced by institutional norms, market expectations, and community attitudes. Whitelaw (2021) emphasises that AI-generated works are often evaluated differently from traditional art, with scrutiny applied to skill, originality, and market value. These dynamics shape decisions about tool adoption, reflecting that technological affordances are inseparable from social and cultural context (Bijker et al., 1987).

Cultural sensitivity is particularly salient. The use of AI in reproducing or reinterpreting stylistic traditions raises ethical questions about representation and appropriation (Crawford & Joler, 2018). Artists’ awareness of these factors affects their engagement with AI, demonstrating that perception is a multidimensional construct, integrating aesthetic, ethical, and social considerations.

6.5 Implications for Art Practice and Education

The findings have practical implications for artistic practice, education, and professional development:

• Practice: Artists can leverage AI to explore new forms, iterate rapidly, and expand creative horizons while retaining agency through selective intervention.
• Education: Integrating AI into curricula encourages experimentation, technical literacy, and critical reflection on ethical, cultural, and creative dimensions.
• Professional Development: Institutions and markets may need to recognise hybrid authorship models and provide guidelines for attribution, copyright, and ethical use.

Educators and policymakers should consider AI as a pedagogical partner, fostering skills that blend human intuition with algorithmic potential. Such approaches encourage reflective creativity and support ethical engagement with emerging technologies.

6.6 Future Research Directions

The study highlights areas for further investigation:

• Comparative analyses across disciplines (painting, digital media, installation) to examine variations in perceptions.
• Longitudinal studies tracking changes in artists’ engagement with AI over time.
• Exploration of cross-cultural differences in perceptions, reflecting global variations in technology adoption, aesthetics, and ethical norms.
• Integration of quantitative measures to assess creativity outcomes alongside qualitative insights.

By addressing these dimensions, future research can provide a more holistic understanding of AI’s role in shaping contemporary artistic practice.

7. Conclusion

This study provides a comprehensive examination of artists’ perceptions of AI-driven tools in visual art production, revealing complex dynamics between technological innovation, creative agency, and sociocultural context. The findings demonstrate that AI tools function not as replacements for human creativity but as collaborative partners that enhance ideation, experimentation, and iterative workflow. Through mechanisms such as generative adversarial networks, style transfer algorithms, and other AI-assisted applications, artists can explore novel visual combinations, hybridise influences, and accelerate creative processes (McCormack et al., 2019; Gatys et al., 2016).

However, the integration of AI is not without challenges. Artists encounter unpredictability in AI outputs, algorithmic biases, and technical barriers, which can influence perceived control and creative agency (Hertzmann, 2018; Crawford & Joler, 2018). The negotiation of authorship emerges as a critical concern, reflecting distributed creativity theory’s emphasis on relational and co-creative processes (Boden, 2016; Bown & Van Meeuwen, 2020). While AI contributes substantially to the generation of novel ideas, ultimate authorship is mediated through human selection, refinement, and interpretation.

Ethical and sociocultural considerations further inform engagement with AI tools. Intellectual property, cultural representation, and environmental sustainability emerge as significant factors shaping artists’ decisions and perceptions (Elgammal et al., 2017; Strubell et al., 2019; Whitelaw, 2021). These dimensions underscore the importance of a responsible and reflective approach to AI integration in creative practice, emphasising that technological capability alone is insufficient without ethical and cultural awareness.

The implications of this study are multi-fold. For artistic practice, AI offers a means to augment creativity and expand conceptual horizons while retaining human agency. In education, AI integration fosters critical reflection, technical literacy, and co-creative competencies. Professionally, institutions, galleries, and markets must consider hybrid authorship, ethical attribution, and culturally sensitive deployment of AI-generated works.

In conclusion, AI-driven tools are reshaping the landscape of visual art production, presenting both opportunities and challenges. Artists navigate these tools in ways that balance innovation with authenticity, technological potential with ethical responsibility, and co-creation with authorship. This research provides insights into these nuanced dynamics, emphasising the transformative role of AI as a creative collaborator in contemporary visual arts.

References

Bijker, W., Hughes, T., & Pinch, T. (1987). The social construction of technological systems. MIT Press.

Boden, M. (2016). Creativity and artificial intelligence. Artificial Intelligence, 85(1-2), 347-356.

Bown, O., & Van Meeuwen, L. (2020). Mapping human and machine creativity in practice. Journal of Creative Behaviour, 54(3), 532-547.

Colton, S. (2012). The painting fool: Stories from building an automated painter. Computational Creativity Research.

Crawford, K., & Joler, V. (2018). Anatomy of an AI system. AI Now Institute.

Davis, H. (2020). Art and algorithms: Exploring AI in art education. Journal of Arts and Technology, 12(4), 211-230.

Elgammal, A., Liu, B., Elhoseiny, M., & Mazzone, M. (2017). CAN: Creative adversarial networks, generating art by learning about styles and deviating from norms. ArXiv.

Gatys, L., Ecker, A., & Bethge, M. (2016). Image style transfer using convolutional neural networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2414-2423.

Gere, C. (2008). Digital culture. Reaktion Books.

Hertzmann, A. (2018). Can computers create art? Arts, 7(2), 18.

Mannan, K.A., & Farhana, K.M. (2026). The Principles of Qur’anic Research Methodology: Deriving the Process of Knowledge from Revelation. KMF Publishers. Open Access (CC BY 4.0). DOI: https://doi.org/10.64907/xkmf.book.pqrm.26.02.12

McCosker, A., & Wilken, R. (2020). Machine vision, generative images and creative practice. Art Journal, 79(1), 94-110.

McCormack, J., Gifford, T., & Hutchings, P. (2019). Ten questions concerning generative computer art. Leonardo, 52(2), 104-116.

Norman, D. (2013). The design of everyday things (Revised and expanded ed.). Basic Books.

Sawyer, R. (2017). Group genius: The creative power of collaboration. Basic Books.

Strubell, E., Ganesh, A., & McCallum, A. (2019). Energy and policy considerations for deep learning in NLP. Proceedings of ACL, 3645-3650.

Whitelaw, M. (2021). The advent of AI art: Implications and artist experiences. Digital Creativity, 32(3-4), 201-219.