This qualitative research study investigates how students at the National Institute of Fashion Technology (NIFT),

Bengaluru engage with Generative Artificial Intelligence (AI) in the context of digital fashion photography

education. As AI tools such as Midjourney, DALL-E 3, Adobe Firefly, and Stable Diffusion are increasingly

integrated into fashion media learning environments, critical questions arise around how students perceive these

technologies, how they influence creative development, and what pedagogical and ethical tensions emerge in

the classroom.

The study employed a qualitative research design, drawing on in-depth semi-structured interviews and open-

ended questionnaires administered to a purposively selected sample of 42 students enrolled in undergraduate and

postgraduate fashion communication and photography programmes at NIFT Bengaluru. Data were collected

structured institutional guidance on the ethical and technical dimensions of AI use; and (iv) cultural and

representational concerns—participants frequently highlighted the inadequacy of AI-generated imagery in

reflecting South Indian and broader Indian aesthetic identities.

The findings underscore the need for AI-integrated, critically reflective fashion photography curricula at NIFT

Bengaluru, and contribute original qualitative evidence to the growing scholarly discourse on AI-mediated

creativity in fashion education.

Development, AI Literacy, Fashion Communication, Visual Aesthetics

increasingly accessible and capable—enabling the production of photorealistic fashion imagery without

traditional photoshoots—students at NIFT Bengaluru find themselves navigating a rapidly shifting creative

landscape that their formal education has only partially addressed.

Generative AI refers to a class of machine learning models—principally diffusion models, Generative

Adversarial Networks (GANs), and transformer-based architectures—that produce original visual, textual, or

multimodal content by learning statistical patterns from vast datasets (Goodfellow et al., 2014; Ho et al., 2020).

In the domain of fashion imagery, tools such as Midjourney v6, Adobe Firefly, DALL-E 3, and Stable Diffusion

XL now enable the generation of highly detailed editorial fashion compositions, garment visualisations, and

styled model imagery. Understanding how student photographers at NIFT Bengaluru perceive, adopt, and

critically engage with these technologies is therefore a matter of immediate pedagogical and professional

lived experiences of learners. Students at NIFT Bengaluru are simultaneously enthusiastic early adopters,

anxious about professional displacement, and under-supported by curricula that have not yet systematically

incorporated AI literacy and critical reflection.

This study seeks to address this gap by exploring the meanings, tensions, and creative processes that NIFT

Bengaluru students associate with generative AI in digital fashion photography, using qualitative methods that

privilege depth of understanding over statistical generalisation.

The academic literature on AI and fashion photography is dominated by quantitative industry surveys and

speculative commentary. Studies specifically examining student learners in Indian fashion institutions are

virtually absent. Furthermore, existing work rarely attends to the cultural specificity of Indian aesthetic contexts,

the findings provide evidence-based guidance for curriculum development and AI integration policy. For

educators, the study surfaces student perspectives that are rarely consulted in institutional decisions about

technology adoption. For the broader fashion education community in India, it offers an original qualitative

account of how the next generation of fashion photographers is constructing their professional identities in an

age of generative AI. For scholars, it contributes to the growing interdisciplinary literature on AI, creativity, and

design education.

photography coursework and personal creative projects?

RQ2: In what ways do students perceive generative AI as expanding or constraining their creative development

as photographers?

RQ3: What tensions, anxieties, or identity conflicts do students associate with the use of AI in fashion image-

making?

RQ4: How do students evaluate the representational adequacy of AI-generated fashion imagery in relation to

Indian cultural and aesthetic contexts?

RQ5: What do students identify as gaps in their current institutional education regarding AI literacy and ethical

use?

foundational taxonomy of computational creativity distinguishes combinatorial, exploratory, and

transformational forms—a framework that remains generative for evaluating what AI can and cannot accomplish

aesthetically. Elgammal et al.’s (2017) Creative Adversarial Network (CAN) experiments demonstrated that AI-

generated artworks could be rated as equivalent in creativity to human-produced works by uninformed

evaluators, a finding with direct implications for fashion image assessment. McCormack et al. (2019) advance a

‘partnership’ model of computational creativity, arguing that AI functions most productively not as a

replacement for human creative agency but as a collaborative extension of it.

In fashion contexts, Rombach et al.’s (2022) work on latent diffusion models—the architecture underpinning

This study draws on two complementary theoretical frameworks. First, Csikszentmihalyi’s (1996) Systems

knowledge and personal disposition), a field (comprising gatekeepers and evaluators), and a domain (the cultural

and symbolic system within which creation occurs). Generative AI disrupts this tripartite model by introducing

a non-human algorithmic agent, and understanding how student photographers at NIFT Bengaluru navigate this

disruption is central to this investigation.

socialisation—a framework that attends to the ways in which learners develop professional self-concepts and

craft identities through their engagement with pedagogical practices and tools. Given that NIFT Bengaluru

students are at a formative stage of professional identity construction, the introduction of disruptive AI tools into

their learning environments raises significant questions about how such identities are being shaped, negotiated,

and defended.

Kim and Cho (2023) found that 64% of fashion consumers in South Korea could not reliably distinguish between

practitioners reported significant ambivalence about the implications for creative authenticity and professional

identity. Their findings—drawn from industry professionals—parallel the concerns anticipated among student

learners in this study.

professionals in India, identifying enthusiasm for AI-generated mood boards alongside resistance to full AI

replacement of photoshoots, particularly on grounds of cultural representation and skin tone accuracy. Their

work foregrounds the cultural specificity of Indian fashion contexts that this study also addresses—but from the

perspective of student learners rather than industry practitioners.

thinking through iterative processes of sketching, prototyping, and peer critique—a framework that now requires

extension to account for AI-assisted ideation phases. More recently, Pham et al. (2023) documented how design

students across disciplines responded to AI tool introduction with a mix of excitement, anxiety, and calls for

greater pedagogical support—a finding that this study anticipates will resonate with the NIFT Bengaluru student

population.

The ethics of AI-generated fashion imagery are addressed by Floridi et al. (2020), who articulate principles of

AI ethics including transparency, fairness, and accountability, and by Pasquale (2020), who raises concerns about

algorithmic opacity and the homogenisation of aesthetic outputs when training datasets are non-representative.

These frameworks inform the study’s attention to cultural and representational concerns.

research with student fashion photographers at an Indian institution examining how they experience, make sense

of, and respond to generative AI in their creative education. Existing work is either quantitative, industry-

focused, or conducted in non-Indian contexts. This study directly and exclusively addresses this gap.

identification. Initial codes were generated inductively from participant data; themes were constructed through

an iterative process of grouping, reviewing, and refining codes in relation to the research questions.

interview participants who reviewed preliminary theme summaries and provided feedback on their adequacy as

representations of expressed experiences. Negative case analysis was conducted to attend to data that did not fit

emerging themes. A reflexivity journal was maintained throughout the research process, and an audit trail of

analytical decisions was documented.

Year 3 B.Des.)

on AI use in assessed work—a finding that suggests this tension is partly produced by pedagogical ambiguity

rather than AI use per se.

(Participant A28, PG Year 2)

This sentiment—that the value of photographic work resides in the evidence of human decision-making and

even imperfection—was common among more experienced students and aligned with theoretical articulations

of creative identity as constituted through process rather than product (McCormack et al., 2019).

institution’s pedagogical preparedness for the AI era. The overwhelming majority of participants (38 of 42,

90.5%) expressed the view that current NIFT Bengaluru curricula did not adequately address AI tools, their

creative applications, their limitations, or their ethical dimensions.

responses, further deepening the pedagogical gap:

than before.” (Participant A34, Year 3 B.Des.)

Students articulated specific desires for structured AI instruction. Their requests included practical training in

AI tool prompting and workflow integration, critical frameworks for evaluating AI-generated output, clear

institutional guidelines on AI use in assessed work, and ethical discussions around authorship, representation,

and professional responsibility.

Postgraduate students, who occupied relatively more autonomous learning positions, expressed a stronger

demand for critical frameworks than for technical instruction—indicating a developmental dimension to

other Indian states—concerned the representational inadequacy of AI-generated fashion imagery in relation to

Indian cultural and aesthetic contexts. Participants consistently identified specific failures of generative AI tools

to produce images that reflected South Indian and broader Indian aesthetics: skin tones, facial features, garment

types, styling traditions, and the aesthetic vocabularies of regional Indian fashion.

Participant A15, a second-year student from Karnataka, described a frustrating encounter:

of the Western-centric datasets on which major generative AI models are trained—a structural bias with direct

implications for the usability of these tools for Indian fashion contexts:

“All these AI tools are trained on Western fashion data. Indian fashion is an afterthought, if it’s included at all.

our aesthetic heritage.” (Participant A38, PG Year 1)

Several students drew connections between AI’s representational failures and broader questions of cultural

power in the fashion industry, expressing concerns that the homogenisation of AI outputs would further

marginalise non-Western aesthetics in a global fashion media landscape already skewed toward European and

North American standards.

Despite these concerns, some participants also identified the representational gap as a site of creative

opportunity—a space where human photographers with embodied cultural knowledge could outperform AI and

assert professional distinctiveness:

“This is where I think AI actually can’t replace us—not yet, maybe not ever. I grew up wearing traditional

Karnataka dress. I know how it moves, how it feels, what it means. No prompt can give an AI that knowledge.

That’s my edge.” (Participant A22, Year 4 B.Des.)

The tension between technological enthusiasm and creative authenticity identified in this study reflects a deeper

process of professional identity formation that is characteristic of advanced design education (Holland, 1997).

Students at NIFT Bengaluru are simultaneously developing technical skills, aesthetic sensibilities, and

professional self-concepts—a tripartite developmental process that AI’s disruption complicates in ways not yet

addressed by institutional frameworks.