INTRODUCTION
2023 marked a watershed moment in the broader adoption of artificial intelligence (AI) technologies, particularly generative artificial intelligence (GenAI). In 2024, GenAI began to be operationalized at scale across numerous organizations, transitioning from experimental tools into business-critical solutions (Singla et al., 2025). Marketing, one of the most dynamic and innovation-driven disciplines, has emerged as a key domain for GenAI deployment.
Industry forecasts signal that GenAI has the potential to increase marketing productivity by as much as 15% of total spending (Cillo & Rubera, 2024). These savings stem from efficiency gains in content production and reduced costs in campaign planning, optimization, and customer engagement (Peres et al., 2023). GenAI, with its ability to autonomously generate text, images, audio, and other media, has already redefined what constitutes a “creative” task.
There are numerous and widely varying studies on AI and GenAI in recent academic literature and industry reports. Quantitative studies have largely taken the focused on large organizations and examined the impact of AI/GenAI on corporate marketing (Singla et al., 2025). Articles more directly focused on marketing itself often rely on in-depth individual interviews, but rarely implement a quantitative survey.
This literature review was conducted in March 2025 utilizing the Scopus database. The following queries were conducted to construct the literature review. The first stage query included “marketing agency(-ies)” or “creative agency(-ies)” and “AI” or “Artificial Intelligence”. As part of stage two, the query was repeated, searching for items related to the phrases "marketing agency(-ies)” or “creative agency(-ies)” and “GenAI” or “Generative Artificial Intelligence.” The results are presented in Table 1. The following queries present the scope of the bibliography searches:
TITLE-ABS-KEY ( ( ( „Marketing agency” OR „Marketing agencies” OR “Creative agency” OR “Creative agencies” ) AND ( “AI” OR “Artificial Intelligence” ) ) )
TITLE-ABS-KEY ( ( ( „Marketing agency” OR „Marketing agencies” OR “Creative agency” OR “Creative agencies” ) AND ( “GenAI” OR “Generative Artificial Intelligence” ) ) )
Table 1.
Bibliometric analysis
| Key words | Marketing agency(-ies) | Creative agency(-ies) |
|---|---|---|
| AI/Artificial Intelligence | 16 | 26 |
| GenAI/Generative Artificial Intelligence | 0 | 3 |
| Total | 3 |
The performed evaluation and selection of the publications obtained from the bibliometric analysis allow us to conclude the existence of a research gap. Wojtkiewicz (2024) elaborates on the use of AI tools to generate images in the context of understanding them as works of art. Moura’s monograph (2024) compiles papers touching on both creativity and AI, viewing them together and following their combined possibilities.
The purpose of this article is to present the results of research on the use of GenAI in the field of marketing creation. Based on the hypothesis that marketing agency employees use GenAI in their work, the following research question was posed: How have marketing agency employees implemented AI into their daily creative processes? A survey was conducted among marketing agency professionals and focused on the identification of key tools, the degree of AI implementation, and its impact on creative processes.
THEORETICAL BACKGROUND
Creative Processes in Marketing Agencies
Creativity is a fundamental component of marketing and advertising, yet it remains a concept with diverse interpretations. Most scholars agree that creativity involves generating novel, valuable, and applicable ideas (Cropley, 2020). Graham Wallas’ (1926) four-stage model distinguishes between the stages of creativity: preparation, the research and construction of the project problem; incubation, an openness to ideas; illumination, the selection of ideas; and verification, the evaluation and development of an idea. The process of project development conducted in marketing agencies can be similarly structured. Sternberg and Lubart (1998) believe that creativity should be associated with the usefulness of the ideas generated, which also favors classifying marketing agency activities as creative ones.
The concept of creativity also appears in the context of “creative industries”. According to the British Department for Digital, Culture, Media, and Sport (1998), these industries include activities that use intellectual property and are driven by individual creativity, skills, and talent that have the potential to create wealth and jobs through creation. The above organization, as well as others, use ambiguous classifications both in terms of nomenclature (creative/cultural sector) and the affiliation of various fields to this group. A systematic classification of selected organisations has been provided in Table 2.
Table 2.
Creativity – world organizations’ classification
| Source | Nomenclature | Contents |
|---|---|---|
| GUS | Cultural and creative industries - areas | Visual arts, audiovisual arts and multimedia, architecture, art education, libraries and archives, advertising, books and press, performing arts, cultural heritage |
| DCMS | Cultural and creative industries | Advertising, antiques market, architecture, crafts, design, fashion, music, film, computer and video games, performing arts, publishing, software, radio and television |
| UNESCO | Cultural areas | Cultural and natural heritage, performances and celebrations, visual arts and crafts, book and press, interactive and audiovisual media, design and creative services |
| UNCTAD | Cultural industries | Visual arts, publishing and press activities, design, creative activities, new media, audiovisual arts, performing arts, traditional cultural expressions, cultural centres |
| Creative Australia | Cultural sector | Aboriginal and Torres Strait Islander arts, languages and cultures; cultural heritage; design; music, performance and celebration, including community cultural development; screen arts, broadcasting and interactive media; visual arts and crafts; writing and publishing activities |
| Creative Scotland | Arts and creative industries | Advertising, architecture, visual arts, crafts, fashion and textiles, design, performing arts, music, photography, film and video, computer games, radio and television, journalism and publishing, cultural heritage, software/e-publishing, cultural education |
| Creative Europe | Cultural and creative sectors | Architecture, cultural heritage, design, literature and publishing, music, performing arts, film and audiovisual industry |
[i] Source: Le Viet-Błaszczyk, 2024
Most of the classifications include the products of marketing activities, and therefore the work of marketing agencies (Le Viet-Błaszczyk, 2024).
Organizations implement different models of marketing processes. They can conduct in-house activities, delegating all or part of the work to in-house teams. Alternatively, organizations can use external resources, such as freelancers or marketing agencies. The latter’s activities cover a wide range of services, from market analysis and strategy planning to advertising campaign creation and optimization. Marketing agencies can operate in different operational models: specialist agencies focusing on a selected area; smaller boutique agencies specializing in niche industries; or agencies offering comprehensive services. There are no clear or rigid rules about what is provided by an agency and what is provided internally within an organization (Hughes & Vafeas, 2019).
Creative processes may appear within the agency’s overall marketing activities — from preparing a strategy for a tender to refining creative lines, creating texts, graphics, audio, video, or publishing daily communications. Creative processes can also assist in drawing conclusions that lead to improving marketing activities. Expectations are rising as marketing is given more and more serious tasks beyond communications. Marketing departments and marketing agencies are tasked with finding new and innovative solutions in terms of pricing, promotion, sales, or cooperation with the environment. Creativity is becoming essential, not only in product or process development, but also in the creation of business models (Garbarski, 2018).
Today’s agencies operate in a dynamic environment, requiring creativity, knowledge of analytical tools, and the ability to adapt to changing trends and technologies. Among these technologies is AI. Recent monographs (Anscomb, 2024) and research articles (Wojtkiewicz, 2024) demonstrate that the use of AI tools is not incompatible with creativity or art. While traditional models view creativity as an inherently individual and human cognitive process, recent literature highlights a paradigm shift toward augmented creativity (Schirrmeister et al., 2026). Consequently, the creative workflow in marketing agencies can be conceptualized as a human-AI co-creation process, where the processes are sociomaterially distributed across human and algorithmic actors.
Artificial Intelligence in Marketing
According to Kotler and Keller (2012), it is crucial to use the latest technologies to find innovative ways to cope with the challenges of the market environment in the 21st century. It is evident that successive media, from print, to radio, to television, to the internet, have been called “new” and have played a significant role in marketing. The most significant revolution for the industry was the invention of personal computers and the Internet (Russel, Lane, 2000). That effectively launched a whole group of new media, i.e., techniques for acquiring, capturing, processing, and transmitting information, data, sounds, and images (Bajdak, 2017). The merits of new media were swiftly identified, including the Internet’s extensive selection, ease of access, convenience, substantial impact, and the facilitation of consumer collaboration. Thereafter, marketing initiatives began to prioritize interactivity, personalization, rapid feedback, co-creation of value, and prediction of customer expectations (Mazurek, 2018). The implementation of sophisticated technological frameworks has given rise to a novel marketing technology, which provides enterprises with a variety of technological solutions (Mazurek, 2018), now including artificial intelligence.
There is a consensus in the literature that changes resulting from the implementation of AI solutions will have a major impact on society, as well as business environments (Leszczynski, Salamon & Zieliński, 2022). Despite this, there is a lack of consensus when it comes to defining AI. For Scherer (2015), it is a machine that makes it possible to perform tasks that would require the human performing them to have specialized knowledge, training, or licensing. Daugherty, Nenterme (2018), and Valin (2018) focus on advanced technology that allows learning, analysis, or problem solving, otherwise understood as performing the tasks of the human body’s cognitive functions. Kaplan and Haenlein (2019) believe that AI allows for the correct interpretation of existing data and learning from the data, as well as using the resulting conclusions to achieve the goals and tasks set for the system through flexible adaptation. AI has begun to be seen as an autonomous entity acting on behalf of humans, enhancing their capabilities to perform new tasks and improve existing ones (Sweet & Narain, 2025).
The use of AI in marketing activities has huge implications. Marketers are forced to manage this change, as it will be an essential skill set in the industry. Automating tasks through AI is expected to allow marketers to focus on generating value for customers and creative thinking (Cannella, 2018). AI can help process advertising agencies’ data from a cost and campaign performance perspective, which can allow for predictions related to agency actions and their impact on achieving better results. For this reason, AI is being used to significantly improve processes in advertising campaigns (Enache, 2020). Agencies are using AI to study lifestyles and consumer behavior and update campaigns in near real time (Tahoun & Taher, 2022). Leszczynski, Salamon, and Zielinski (2022) surveyed the use of AI in marketing agencies. In their study, respondents indicated that they use AI for repetitive and analytical activities, while they are skeptical about using the technology in everyday duties. However, respondents believe it could be of great importance in the future.
One of the most transformative branches of AI for creative industries is generative AI (GenAI). Unlike traditional AI, which analyzes data to derive insights, GenAI can create entirely new artifacts: images, texts, audio, video, and more (Kanbach, Heiduk, Blueher, Schreiter & Lahmann, 2023). This shift from an analytical to generative capacity fundamentally alters the creative value chain.
Tools like ChatGPT, Midjourney, and DALL-E exemplify the power of GenAI.Midjourney, for example, offers an image generator using typed commands (“prompts”), and ChatGPT is dedicated to generating text, translating it, and answering questions by presenting specific knowledge. GenAI has the potential to improve the efficiency of marketing content and reduce the cost of marketing activities. It has been shown that some of the tasks of marketing employees using GenAI take less time, and their performance is improved (Noy & Zhang, 2023; Peres et al., 2023CITATION NEEDED). This technology reduces barriers to entry for performing creative processes (Noy & Zhang, 2023).
A dynamic area of AI research is also computational creativity, which allows algorithms to create music or write texts (Galloway & Swiatek, 2018). According to Bouschery et al. (2024), coupling creatives with generative systems provides cognitive stimulation that bypasses traditional group brainstorming limitations, such as production blocking, by activating semantic memory nodes on demand.
Marketers are using GenAI to create advertising campaigns. For example, in Poland, marketers used GenAI to create a campaign for Adamed Pharma. The company’s press office said that by using such a procedure, production costs were significantly reduced (Drynko, 2022). The Spanish store chain Mango used generated avatars and had AI help design clothes and provided inspiration for fabrics (Wirtualne Media, 2024).
Academic and industry opinion remains divided on AI and creativity. While some scholars maintain that creativity remains a uniquely human domain (Cannella, 2018), others argue that computational creativity is both possible and increasingly evident in practice (Galloway & Swiatek, 2018). It is within this debate that the current study positions itself, exploring how professionals in marketing agencies perceive and utilize GenAI in their daily creative work.
RESEARCH METHODOLOGY
In order to achieve the research objective, the following hypothesis was set: Marketing agency employees use GenAI at work. In conjunction, we refined the main research question: How have marketing agency employees implemented GenAI into their daily creative processes? This was supported by five auxiliary questions:
How often do they use GenAI at work?
What is their opinion on this use of GenAI in creative work?
For which creative work do they use GenAI?
What are the most popular tools?
What attitudes toward GenAI are there in the agency environment?
The research design was structured as a sequential explanatory study, combining an initial theoretical phase with an empirical quantitative phase. The rationale for this approach was twofold: first, to build a theoretical framework and ensure the research questions addressed relevant gaps identified in the literature; second, to measure actual GenAI usage and perceptions among professionals engaged in creative marketing work.
The first stage involved synthesizing and analyzing available academic and industry literature. This included articles published in peer-reviewed journals and reports. The analysis focused on identifying (1) the main categories of GenAI tools in use, (2) the scope and patterns of their implementation, and (3) the attitudes of employees and organizational stakeholders towards AI-supported creativity. This stage was instrumental in constructing the survey questionnaire, ensuring content validity and relevance of the items to current agency realities.
In the second stage, a quantitative study was conducted using the Computer-Assisted Web Interview method, with a self-administered online questionnaire designed in Google Forms. This method was chosen due to its accessibility, speed of distribution, and the ability to reach geographically dispersed respondents while maintaining anonymity and comfort of response (Matejun, 2020). The questionnaire consisted of 12 closed-ended questions (single-choice, multiple-choice, and Likert-scale items), as well as one open-ended question designed to elicit qualitative reflections on GenAI use in creative tasks.
The survey was conducted in February and March 2025. The sampling method was purposive (non-probability sampling) and focused on professionals employed in marketing or creative agencies. To mitigate selection bias and ensure high-quality, industry-specific insights, a dual-channel digital recruitment strategy was deployed through professional networks. The acquisition of respondents encompassed verified professional networks of marketing practitioners operating in Polish agencies, complemented by recruitment via dedicated corporate communication platforms of global marketing entities within an established communications group. This purposive approach combined targeted selection with elements of virtual snowball sampling. Consequently, the sample was deliberately oriented toward innovative, digitally mature agency ecosystems and professional cohorts structurally predisposed to acting as early adopters of emerging technologies.
To verify the appropriateness of the sample, an introductory question was included asking whether the respondent currently works in a marketing agency; only affirmative answers were retained in the final analysis. The final sample consisted of 84 completed surveys, of which 79 were validated as being from agency employees. The demographic structure was analyzed in terms of age and company size, providing context for the interpretation of the results.
To ensure the reliability and usability of the survey, the questionnaire was piloted among a small group (n=6) of agency professionals and academic reviewers prior to its distribution. Based on their feedback, minor adjustments were made to improve the clarity and logical flow of the questions. The survey achieved a relatively high completion rate, suggesting that the topic was engaging and relevant for the target group.
Data were analyzed using descriptive and inferential statistics. The Likert-scale responses were summarized using means and standard deviations. To test the relationship between GenAI usage frequency and perceptions of its utility or risks, Spearman’s rank correlation coefficient was employed due to the ordinal nature of the data. Additionally, the Kruskal-Wallis H test was used to compare attitudes across different usage-frequency groups. For the qualitative insights derived from the open-ended question, we conducted a thematic analysis guided by a six-phase framework (Ahmed et al., 2025). The analysis began with data familiarization through immersive reading of the qualitative responses. We then generated initial semantic and latent codes to label key features related to GenAI usage. These codes were clustered into broader candidate themes, which were iteratively refined to ensure internal homogeneity and external heterogeneity. Finally, the themes were defined, named, and integrated into a coherent narrative report to provide qualitative depth to the quantitative findings.
In sum, this two-stage mixed-method design provided both theoretical grounding and empirical insight into how GenAI is reshaping creativity within marketing agencies. The methodology was tailored to capture both observable behaviors (tool usage) and underlying perceptions (attitudes, concerns, optimism), offering a comprehensive look into this emerging area of practice.
RESULTS
A total of 84 respondents participated in the survey, of whom 79 declared themselves to be employees of marketing agencies, and this subsample was used as the basis for the analysis. The demographic breakdown, as presented in Figures 1 and 2, shows a predominance of participants aged 25–40 (65.8%) and employees working in medium-sized agencies with 50 to 249 people (67.1%).
Figure 1.
Respondents’ age
Source: own study
Figure 2.
Company size
Source: own study
As many as 97% of respondents declared that they use GenAI at work. Among them, 41.6% use the technology daily, 32.5% use it 2–3 times a week, 14.3% use it once a week, 7.8% use it a minimum of once a month, and 3.9% use it less frequently. 76.6% confirmed that they have used GenAI more often in the past 6 months than before, and 66.8% planned to use it more often in 2025 than the previous year.
Agency employees use the technology in a variety of areas, shown in Figure 3. Although the study primarily focused on structured response categories, individual free-text entries were also taken into account. Among these single, open-ended responses grouped under the ‘Other’ category, respondents highlighted specific tasks such as search engine optimization/advertising, taking meeting reports, edits, analyzing customer feedback, website building, and occasionally to replace Google Search.
Figure 3.
GenAI usage
Source: own study
The most common areas of work involving GenAI include writing marketing copy, visual design (e.g., ad creatives, banners), social media content generation, trend analysis, and idea prototyping. Tools mentioned in the responses include ChatGPT, Google Gemini, MidJourney, Adobe AI, Canva AI, and DALL·E. Respondents also cited specialized applications like DeepL Write, Grammarly, Surfer SEO, Synthesia, Runway ML, and newer entrants such as Claude, Recraft, or Suno. This demonstrates not only the breadth of use but also a certain technological literacy and exploratory mindset among users.
An interesting aspect in the context of the rise in popularity of GenAI is employees’ perception of this technology’s impact on creative activities. Respondents rated their stance towards the use of GenAI on a five-point Likert scale (1 - “strongly disagree”, 5 - “strongly agree”). A summary of the average ratings for part of the responses is shown in Table 3.
Table 3.
Opinions on GenAI in creative marketing activities
| What’s your opinion about GenAI in creative work? | Average rate |
|---|---|
| It has high potential | 4.36 |
| It fosters efficiency | 4.26 |
| It can be inspiring | 4.25 |
| It fosters innovation | 4.03 |
| I’m worried about copyright infringement of GenAI | 3.48 |
| I consider myself AI literate | 3.21 |
| GenAI destroys creativity | 2.83 |
| I’m afraid that GenAI will take away my job | 2.23 |
The highest-scoring claim was “Gen AI has high potential” (M = 4.36), indicating a dominant belief in the value of these tools in the context of creative work. The assertions “GenAI fosters efficiency” (M = 4.26) and “GenAI can be inspiring” (M = 4.25) both scored very high, suggesting that respondents recognize both the creative and practical aspects of using this technology. Regarding the impact on creativity, most respondents do not see AI as a threat to creativity (M = 2.83). Rather, concerns focus on copyright infringement (M = 3.48). Finally, it is worth noting that the self-assessment of competence in GenAI received a score of M = 3.21, indicating a moderate level of confidence in using these technologies.
To examine whether the frequency of use of GenAI is related to opinions about the technology, a Spearman correlation analysis was conducted. Values for frequency of use were numerically coded on a scale from 0 to 5, where 0 – Never and 5 – Every day. Respondents’ opinions about GenAI were converted to numerical values according to a five-point Likert scale. A correlation test showed significant correlations between the frequency of GenAI use and several aspects of perceptions of the technology. The results of the analysis are shown in Table 4.
Table 4.
Relationship between frequency of GenAI use and opinions about it
| Opinion | Spearman corelation | p-value | Kruskal-Wallis test statistic | p-value |
|---|---|---|---|---|
| It fosters efficiency | 0.3924 | 0.0044 | 8.0095 | 0.0458 |
| I consider myself AI literate | 0.4554 | 0.0007 | 12.6222 | 0.0055 |
The strongest correlation occurred for self-assessment of AI competence (ρ = 0.46), which may suggest that the more often a person uses AI, the more confident he or she is with the technology. In addition, positive attitudes toward AI, such as perceiving it as an efficiency-enhancing tool (ρ = 0.39), also increased with frequency of use. In order to test for differences between the several groups, the Kruskal-Wallis H test was applied—the results are shown above. The results indicate that differences between groups exist and are at a statistically significant level.
In conclusion, the results suggest that regular use of AI promotes a better understanding of the technology and reinforces the belief that it has a positive impact on creative processes.
Respondents also answered questions about attitudes toward AI in and around the agency. The results are shown in Figure 4.
Figure 4.
Attitudes toward GenAI in marketing
Source: own study
To analyze the relationship between the frequency of GenAI use and attitudes toward it—positive, negative, or neutral—a Spearman test was conducted. Spearman’s correlation coefficient was 0.38, indicating a moderate positive correlation, suggesting that the more often someone uses GenAI, the more positive their attitude toward it. The p-value (0.0054) is statistically significant (p < 0.05), suggesting that the relationship between these variables is not random. A Kruskal-Wallis H test was also conducted, confirming that there are statistically significant differences in attitudes toward GenAI between groups with different frequencies of use (H-statistic = 9.84, p-value = 0.01999 < 0.05).
The questionnaire also collected 23 detailed opinions in response to an open-ended question on how to use GenAI in creative work. Respondents mainly indicated the ability to generate graphics (including moodboards), edit images, and fill in missing elements in visualizations. Respondents pointed to the common use of tools such as “generative fill” in Photoshop, Stable Diffusion, and Adobe Firefly. Regarding written text, responded pointed to GenAI’s assistance in creating and optimizing marketing content, analyzing documents, and searching for inspiration. Users also use GenAI for research, trend analysis, and data verification. Many note that it streamlines work and helps organize work. Some respondents also note the negative aspects of AI, pointing to its impact on the labor market, copyright issues, and the environment. They emphasize that the generated content needs to be verified and can lead to a decline in creativity. Despite these concerns, respondents acknowledged that these tools could support daily work.
From the entirety of the research conducted, it can be concluded that marketing agencies have indeed introduced GenAI into their daily creative work. On the one hand, the results indicate the usage of several tools with this technology and confirm the perception of GenAI’s high potential, especially in terms of the efficiency of creative processes. On the other hand, the results suggest the possibility of a competency gap in AI and copyright concerns. Findings on the relationship between the frequency of GenAI use and attitudes toward GenAI may suggest that more frequent use promotes a perceived increase in the effectiveness of creative work. In addition, it is inferred that more frequent use of GenAI naturally results in employees becoming more familiar with the technology, making them feel confident in using it. Thus, it seems that the popularity of generative artificial intelligence is somehow related to the increasing implementation of this technology in agencies and the positive attitude towards it. It is recommended that for more effective adoption of GenAI in marketing, the focus should be on education, training, and knowledge sharing in and around an agency. It is also worth raising topics of concern to introduce appropriate remedies and engage in constructive discussion.
Theoretical Implications
First, the results challenge traditional linear models of creativity, such as the Wallas (1926) four-stage model, by introducing a new actor into the process: the generative algorithm. Whereas classical theories position creativity as an inherently human sequence of ideation and evaluation, GenAI tools intervene not merely as aids, but as co-creators. This supports emerging views in the literature on augmented creativity, or algorithmic creativity, which conceptualize the creative process as a hybrid of human intent and machine output (Galloway & Swiatek, 2018; Noy & Zhang, 2023).
Second, the research demonstrates that exposure to and use of GenAI enhances perceived creative self-efficacy (Reich & Teeny, 2025). Frequent users of GenAI reported higher confidence in their abilities to apply the technology meaningfully (Pellas, 2023; Zhang & Xu, 2024). Similar observations have been made in earlier studies, where individuals who engaged regularly with GenAI tools reported increased confidence in navigating creative tasks, especially in text generation and visual design (Noy & Zhang, 2023; Peres et al., 2023). The implication is that GenAI provides practical outputs and supports user development by enhancing their sense of competence and expanding their creative repertoire (Kanbach et al., 2024; Cillo & Rubera, 2025).
Third, the study indicates a need to revisit the boundaries between tool and creator. Respondents described GenAI as both assistant and ideator, blurring the line between operator and author. This raises theoretical questions related to intellectual property, authorship, and accountability in creative industries. It also suggests that the locus of creativity is shifting from the individual to a distributed model, involving human-AI interaction, where meaning is co-constructed.
Moreover, these findings support theories that position creativity not solely as a trait or process, but as contextually situated within organizational structures and technologies. In this view, tools like GenAI are not neutral instruments but shape creative practices by affording new possibilities and constraints. The adoption of GenAI within agencies may thus reflect a broader evolution in the socio-technical systems that underpin creative work.
Finally, the study provides preliminary evidence that GenAI adoption correlates with organizational culture and maturity. Respondents from larger agencies often reported structured implementation, clearer guidelines, and formal training processes. This may inform theoretical models of technology assimilation in marketing organizations, where culture, size, and leadership influence how innovation is embedded in daily operations.
In sum, this study invites scholars to further theorize the blended nature of creativity in AI-supported environments. Rather than displacing human ingenuity, GenAI appears to reframe it—offering new lenses through which creativity, agency, and authorship can be understood.
DISCUSSION
Every study is associated with certain limitations. In this case, it must be acknowledged that the chosen survey technique involves self-declaration. Another challenge that researchers often face is the reluctance of respondents to participate in surveys (Matejun, 2020). Due to the small research sample, the authors treat this study as a pilot. Thus, expanding the number of respondents may lead to a broader and more in-depth look at the topic, with additional uses of GenAI addressed in the research questions. The authors plan to enlarge the sample soon, to expand statistical analyses of the relationship between other demographics and opinions, tools, and areas of GenAI use.
Leszczynski, Salamon and Zielinski (2022) also surveyed the use of artificial intelligence in marketing agencies, and some respondents were skeptical about using AI technology in everyday duties. Their study differs slightly from the outcome of this research. Here, the results indicate a moderate self-assessment of familiarity with GenAI. However, it should be noted here that this technology is considered new, and familiarity with it may improve over time. Nonetheless, it is worthwhile to delve deeper into the topic and conduct qualitative research to learn about the causes of a given situation, as well as to discuss possible solutions.
The findings suggest a significant shift in how creative work is conceptualized and executed within marketing agencies. The high percentage of GenAI adoption (97%) and its frequent use indicate not just technological experimentation but a reconfiguration of creative routines. This supports the growing view that GenAI tools are not marginal add-ons but are being absorbed into the core of creative workflows (Peres et al., 2023; Kanbach et al., 2024).
Another noteworthy observation is the alignment between usage frequency and positive attitudes toward GenAI. Employees who use GenAI more often tend to evaluate it more favorably, both in terms of its usefulness and their own digital competence. This dynamic suggests a feedback loop, where increased usage enhances familiarity, and familiarity leads to reduced skepticism—a pattern confirmed in similar studies (Noy & Zhang, 2023). In organizational practice, this highlights the importance of facilitating access to GenAI tools and offering learning opportunities that enable employees to build experience and confidence.
The data also reveal heterogeneous adoption patterns depending on organizational size. Smaller agencies appear more agile in adopting GenAI for diverse creative purposes, while larger organizations report more structured, policy-driven implementations. This may reflect broader differences in risk tolerance, infrastructure, and governance. It also reinforces the argument that organizational context plays a key role in how GenAI technologies are integrated, which resonates with the socio-technical systems perspective in innovation adoption (Mazurek, 2018).
Although most respondents rejected the notion that GenAI “destroys creativity,” some expressed nuanced concerns about the quality and originality of generated outputs, the need for human oversight, and ethical implications (Wojtkiewicz, 2024). These views point to the complex nature of AI adoption in creative domains—where the promise of efficiency and inspiration coexists with legitimate questions about authorship, ownership, and creative integrity.
Finally, the open-ended responses add richness to the quantitative findings by highlighting how employees actually use GenAI tools. Their examples illustrate practical utility—such as creating moodboards, optimizing texts, and organizing ideas, while revealing a reflexive awareness of GenAI’s limitations. Many respondents treat GenAI as a starting point or assistant, rather than a final creative authority. This behavior suggests that professionals are negotiating a new creative partnership between human and machine, one that requires both technical skill and critical judgment.
In sum, the results of this study portray marketing agencies as experimental and adaptive environments, where the integration of GenAI is progressing rapidly, but not without reflection. Future research should explore how these tools evolve from experimental support into strategic assets, and how creative professionals balance automation with originality in the long run.
These findings reinforce the view that GenAI is not only a technological innovation, but a catalyst for rethinking creative work structures and knowledge processes within marketing organizations. While computer science and business literature have thoroughly examined the technical aspects of GenAI, its practical impact on the day-to-day routines of creative professionals remains underexplored. This study contributes to bridging that gap by offering an empirical perspective on how GenAI is already reshaping creative decision-making and team dynamics in marketing agencies.
CONCLUSIONS
A survey on the use of generative artificial intelligence (GenAI) in marketing agencies’ creative processes showed that the technology has been widely adopted in daily work. As many as 97% of respondents confirmed using GenAI at work, and nearly three-quarters use it at least weekly, with over 40% using it daily, indicating that GenAI has become a regular element of creative workflows. Therefore, the research hypothesis cannot be rejected. The following paragraphs synthesize the empirical answers to the five research questions posed in the study, addressing the frequency, perception, application areas, tool preferences, and organizational attitudes toward GenAI use in creative work.
In terms of application, GenAI is used primarily for text generation, image creation, concept development, and research support. Respondents noted using it to generate marketing copy, social media posts, moodboards, visuals, and inspiration for strategic ideas. The most frequently mentioned tools were ChatGPT, Google Gemini, MidJourney, Adobe AI, and Canva AI, with many other niche applications also listed, pointing to a technologically literate user base.
Surveyed professionals expressed a predominantly positive opinion about GenAI. They rated its potential and efficiency-enhancing capabilities highly (M > 4.2) and recognized its role in supporting creativity and innovation. A minority viewed GenAI as a threat to creativity (M = 2.83), and even fewer feared job loss. Concerns centered on copyright infringement and the need for content verification. The average self-assessed AI competence was moderate (M = 3.21), suggesting there is room for further development.
The prevailing attitudes within agency environments were generally open and optimistic. The frequency of GenAI use correlated positively with both individual confidence and organizational acceptance, suggesting that early adopters may help shape cultural norms around AI use. Open-ended responses confirmed that employees treat GenAI as a supportive tool—valuable in ideation and execution, but not a replacement for human insight or critical thinking.
The results also suggest that GenAI adoption is not uniform: differences emerge across agency size, technological maturity, and internal policies. This heterogeneity indicates that successful integration of GenAI into agency workflows depends not only on access to tools, but also on organizational culture and leadership support.
From a managerial perspective, agency leaders should not treat GenAI merely as a cost-cutting automation tool. Instead, its implementation requires a strategic approach—one that combines training, transparent communication, and a clear articulation of ethical boundaries. Investing in employee upskilling and encouraging experimentation may help maximize GenAI’s creative potential while fostering trust and engagement within teams.
On the theoretical level, the study provides evidence that GenAI transforms the structure of creative processes, shifting them from linear, human-centered models to hybrid, co-creative systems involving algorithmic output. Furthermore, the blurred boundary between tool and author points to a broader redefinition of authorship and originality in marketing creativity, which should be further explored in future conceptual work.
Future studies should expand the sample to further analyze the impact and understanding of trends within the influence of GenAI on selected aspects of creativity. The research be supplemented with in-depth individual interviews to better understand the attitudes and concerns of marketing agency employees toward GenAI. An interesting direction for future research would also be an analysis of the long-term effects of artificial intelligence adoption, as well as correlational studies of actual, rather than declarative, data on GenAI use.