Generative AI (GenAI) tools are becoming increasingly prevalent in various industries, including the gaming industry. Some consider GenAI as the future of game development, providing opportunities for improved efficiency and automation of several aspects such as artwork generation, programming, and storytelling. Others are equally concerned with potential risks, including concerns regarding copyright, employment, and the lack of qualitative output. However, research on game industry professionals' perceptions and use of GenAI for game development is scarce, focusing on specific target groups (e.g., student and indie developers) or specific GenAI tools (e.g., image generation tools). The current paper includes exploratory in-depth interviews with 20 Belgian game developers to determine their attitudes and usage of GenAI. While some developers were cautiously optimistic about its future potential, most developers were rather skeptical about GenAI's usefulness. Developers agreed that GenAI was most useful for coding and other technical tasks over creative aspects, as a supporting tool. Moreover, use of GenAI mostly occurred during early development, drawing inspiration for name giving of in-game objects and environments, creating concept art for mood boards, or use GenAI voice lines for early prototypes. Quality concerns were mentioned most frequently, as developers criticized GenAI tools' mediocre storytelling quality, inefficient image generation, but also for providing buggy coding solutions. Employment loss was also a persistent concern, while ethical concerns regarding copyright and IP were less prevalent. Interestingly, they were also concerned about how players' interactions with GenAIcreated content could lead to inappropriate situations that were out of their control. These findings are discussed within current knowledge on the topic, discussing future research opportunities and implications for game developers, educators, and policymakers.
Generative AI (GenAI) is becoming an indispensable part of daily life. It has the potential to
promote efficiency and innovation in various industries, including the gaming industry. GenAI can
automatize modelling of in-game objects and characters, or provide realistic voice lines and
automatically generate NPC dialogues [
Despite GenAI’s recent surge in capabilities and popularity (for game development), research on
the use of and attitudes toward GenAI for game development is still scarce [
Traditional AI—one of the core aspects of games, in which predetermined mechanisms and
algorithms ensure that non-player characters (NPCs) and the game world itself respond logically to
player input [7]—is seemingly becoming outdated with the rise of GenAI. This perhaps bold claim is
supported on several strengths and opportunities this emergent technology can offer game
developers. GenAI has the potential to automate several processes allowing developers to focus on
more creative and important aspects. This automation occurs, for example, through GenAI-powered
procedural generation of new and changing game environments [
Additionally, GenAI can provide aid during the early phases of the development process [
Yet another opportunity for the integration of GenAI in game development processes lies in the
promotion stage of the game [
A fourth and final strength, situated on a more abstract level, is that GenAI’s integration in game
development can lead to a more cost-effective game development [
Important to note here is that these strengths and opportunities of GenAI tools for game
development have been identified by researchers as context-dependent. For one, the usefulness of
GenAI tools differs significantly between artistic and technical work-related processes. Research has
consistently shown that the introduction of GenAI has made life easier for programmers. The
acceptance of GenAI tools for programming is high [
While some consider GenAI to be the biggest innovation driver in the gaming industry since virtual and augmented reality technology [23], it's not only (GenAI-generated) sunshine and rainbows. There are at least four major downsides and potential risks to consider regarding GenAI-powered game design. First, there are some general concerns regarding GenAI’s environmental impact, as these large language models (LLMs) require tremendous amounts of computational power in energyconsuming servers [24]. Given the ongoing debates about making the gaming industry more ‘green’ [25], the growing reliance on GenAI tools could pose potential ecological concerns [26].
Second, the creation of game narratives, characters, dialogues and entire game worlds through
GenAI tools leads to several ethical concerns. One of the main challenges that developers identified
are the potential copyright issues that accompany the usage of GenAI tools [
Third, career concerns relate to the consistent fear that increased usage of GenAI tools will
eventually lead to less employment opportunities among voice actors, writers, visual artists, and
even game developers [27]. A recent survey among game developers at the Game Developers
Conference [
This fear of employment loss (and transformation) exists despite the notion that GenAI tools are,
at least at this moment, not there yet. For one, they still need human input to create assets, coding or
artwork. GenAI is not able to “replace the essential human creativity, imagination and artistry that
goes into developing new video games” (para. 8) [8] just yet. Furthermore, they still produce many
errors and overall lack basic quality as well as efficiency within existing workflows. These aspects
can be considered as the fourth main risk factor for GenAI’s integration in game development:
practical concerns. Perhaps the most often mentioned concern here is the lack of consistency when
creating artwork through GenAI tools [
Overall, these weaknesses and risks surrounding the usage of GenAI tools for game development
concern developers, preventing them from implementing these tools in their daily workflow. A clear
example of this comes from Boucher et al. [
Existing literature on generative AI for game development revealed that developers see the most
opportunities for the integration of GenAI tools in early game development and for technical tasks
(i.e., coding, programming [
The few studies that have focused on game developers’ perception and use of GenAI either
focused on specific target groups—student and early career developers [
RQ1: Which strengths and opportunities do Belgian game developers identify regarding the use of GenAI tools for game development? RQ2: Which weaknesses and risks/concerns do Belgian game developers identify regarding the use of GenAI tools for game development?
This study employs a qualitative research design to examine how game developers in the Belgian game industry perceive and use generative AI. Data were collected through in-depth interviews, with a semi-structured questionnaire allowing for consistency and flexibility across respondents [30]. Compared to quantitative approaches, this qualitative approach allowed for a much deeper and more detailed underlying reasoning from the industry on how they perceive and use GenAI for game development.
Before data collection started, respondents received an informed consent form briefing them regarding the use of audio recording during the interview, assurances of confidentiality and anonymity of the interview, and their right to withdraw from the study at any given time.
The interviews were part of a larger project focused on the intentions, strategies and techniques game developers have and use to evoke eudaimonic player experiences (EUPX), or experiences that foster personal growth and help players realize their full human potential [31], which includes emotionally moving, (self-)reflective, and socially connecting player experiences [32]. Besides asking about their eudaimonic intentions and development strategies, the interviews also inquired about recent innovations and technological advancements in the game industry, such as generative AI. Questions for this particular topic related to the use of GenAI in general and toward the design of games that provoke EUPX in particular.
Respondents were recruited using a combination of convenience sampling (i.e., through the author’s personal network of game developers) and snowball sampling (i.e., recruiting developers from the network of respondents that were already recruited). The sample of this study consisted of 20 Belgian game developers. 16 men, two women and two non-binary people were interviewed, with a mean age of 32 and an average 5.65 years of experience in the gaming industry. The study included both established members in the Belgian game industry and six game design students, interns, or developers who just started their first game (studio). Further diversification was made in the different functions someone can have within a game studio, including CEO’s, founders, or game directors; artists and creative or art directors; programmers or tools developers; writers/scripters working on the narrative; and sound designers. An overview of the respondents’ background can be found in Table 1 (Appendix A).
The interviews were transcribed verbatim and coded through NVivo 1.7.2. [33] to organize, structure and analyze the interview transcripts. Following a thematic analysis approach [34], the transcripts were first coded openly and inductively—codes were derived from the data itself—to structure and reduce the data into clear text codes. Conceptually similar codes were then put into overarching codes or themes, merging codes when they were conceptually identical. After reviewing and renaming these themes, the final phase of analysis included connecting the identified themes associated with GenAI to the posed research questions and reporting the findings in the next section.
The thematic analysis resulted into two smaller themes (i.e., ‘GenAI and emotion/eudaimonia’ and ‘Specific cases and examples of GenAI use’) and three major themes: ‘General attitudes toward GenAI,’ ‘Advantages and uses of GenAI,’ and ‘Weaknesses of and concerns toward GenAI’. These larger themes can be further broken down into 16 subthemes, with three subthemes discussing general attitudes (e.g., ‘GenAI promising evolution’ and ‘Against AI use’), eight subthemes addressing strengths and opportunities (e.g., ‘AI images,’ ‘AI for executive/technical tasks’ and ‘AI as supporting tool’), and five subthemes addressing concerns and risks (e.g., ‘Concerns about quality GenAI’ and ‘Concerns about employment’). Since this paper does not specifically focus on the emotional and eudaimonic potential of integrating GenAI into game development processes, and the results section will provide concrete examples of how respondents use GenAI tools, we will concentrate on the three main themes and their corresponding subthemes.
The first theme gathers developers’ overall attitudes toward GenAI for game development. Some were quite positive about this, calling it “the big trend” of game development (Intw. 12). Others were cautiously optimistic. One respondent mentioned that it is “complicated too because, as an artist, [generative] AI is a subject that’s on everyone’s lips. But I’m not 100% against [generative] AI, to be honest, like it can help in some ways” (Intw. 18). Another developer stated that, while he currently did not use GenAI that much, he suspected that it is something to keep an eye out for in the future: “I'm almost certain that with [generative] AI, if that is worked out a little more, that that [providing game designers with more options to do something cool] can be done” (Intw. 11). Yet others were more neutral toward GenAI, as they were not yet convinced of its potential: “We are at the beginning of [generative] AI. So, who knows what’s out there in five years of ten years, but right now, I don’t see the added value yet for our sector, our industry.” (Intw. 1) However, and perhaps surprisingly, most voices regarding GenAI tools were rather skeptical or downright negative of their inclusion in game development processes. Some mention the contextspecific strength of GenAI tools: “There are certainly useful things about it, but in the creative aspects I am more of a humanist” (Intw. 13). Others repeated concerns connected to employment risks—“We say no, sorry, we cannot do that because that was done with generative AI and we are now taking work away from graphic designers and illustrators” (Intw. 10)—and yet others mentioned more qualityrelated concerns of GenAI as creating less original and creative content: “You also see that there are many games that lose a lot of heart and passion because they start using [generative] AI a lot” (Intw. 3). One respondent voiced a clear negative attitude toward GenAI tools, based on ethical concerns regarding copyright issues: “I am very, very, very, very hard against it. Generative AI is one of the worst inventions of the last 50 years. (…) Actually making something from scratch based on a prompt (…). Yes, that is just theft. That is theft.” (Intw. 9)
In this second theme, developers’ perceived strengths and identified opportunities surrounding the use of GenAI can be divided into two major categories: strengths and opportunities that relate to specific game elements or stages of the development process, and more overarching opportunities for the game development process as a whole.
In terms of specific strengths, some respondents mentioned they used GenAI tools to draw inspiration for their game, for example, when needing to come up with names for certain elements within their game. For their factory automation game, one developer used GenAI to ask “Now give me 200 names for [factories]. (…) The titles of our factories, you can still have AI extract [those] now” (Intw. 1). Another respondent mentioned a similar use of GenAI tools: “Names of places are tricky. So, coming up with place names that are truly unique, some people give (…) Then you give a description of your area and then ChatGPT [gives you] a list of names that you can then either continue to create from, namely by linking that list together, or by just using one of them.” (Intw. 16)
Next to name giving as a use for GenAI tools, one respondent (Intw. 16) mentioned that he knew a game company that used GenAI-generated art for the background images (e.g., a city, village, or forest) of certain cards designed for a board game. However, he was one of the few developers to be convinced of GenAI’s usefulness to create game environments and other artwork: “Yes, if it’s really an abstract background. Usually there’s more to it. There are different layers work together” (Intw. 6), signaling the complexity of working with GenAI tools as (environment or character) artists for games. This sentiment reflects the strong consensus among the interviewed developers that GenAI tools hold greater potential for supporting executive and technical tasks in the development process, rather than creative ones. Providing aid with programming and coding tasks is the number one strength developers mentioned: “For a coding problem, if there’s something where we totally do not know how to get started on that, it’s helpful for getting a synopsis or something like that. Or a place to start looking” (Intw. 11). One respondent (Intw. 17) mentioned using ChatGPT as a tool to create games, which he argued sometimes helped him with coding tasks, but no other tools were specifically mentioned. Furthermore, the argument of GenAI tools being able to speed up some processes was made by another respondent: “if it's to speed up some processes, making things easier for everyone, make the [generative] AI conduct some tasks that are not fun at all. That's important” (Intw. 18).
Another subtheme that several respondents addressed is the impact of GenAI on a game’s narrative and players’ agency within both that narrative and the larger game world. Some respondents saw the potential of this, stating that integrating games with LLMs can create “a sort of player freedom, (…) there’s a fascinating side to that” (Intw. 3) and offer a more dynamic experience where players can provide input. This is, however, still in a hypothetical stage, as none of the respondents mentioned they were working on such a game project. They rather talked about this as a future opportunity for game developers: “[Generative] AI could possibly take over parts of the game. Games could become bigger, and could also become more reactive” (Intw. 14). Another respondent talked about this more in detail: “The big trend of course is [generative] AI. (…) By being able to better track what your player is doing, you will also be able to respond to that player better and actually craft a very unique path for your player. (…) What everyone is now full of a Baldur's Gate or a Kingdom Deliverance, where you can make so many choices, have so much agency. That will only grow exponentially in the future. That agency. And then especially in the narrative.” (Intw. 12)
As a final specific strength, some respondents mentioned that audio elements in games could also benefit from the integration of GenAI tools. However, similar to the narrative elements, they rather see this as a future innovation than something that is already happening right now. One respondent further reflected on this: “you feel that you [can] have a smart audio system that allows you to adjust your audio to what you want your players to feel. Then, you can go very deep in terms of emotion” (Intw. 12).
Switching gears to more overarching opportunities developers perceived, the biggest argument a lot of the interviewees mentioned was that GenAI tools are very useful during early development stages. Several developers mentioned they use GenAI tools to “get some quick ideas without having to spend too much time” (Intw. 19). Using these tools in the concept phase, for instance, to visualize your thoughts and present them to other members of the development team, is frequently mentioned: “Conceptually, if you're really thinking about having a concept that you can give to an artist. This is kind of a mood board. This is something we want to create. For that mood board, you can use AIgenerated images” (Intw. 6). Next to using these tools for mood boards and conceptual art, students game design also used GenAI to draw inspiration from, as one respondent mentioned: “They use AI images to get inspiration for their projects. They had to create a murderer, for example, (…) for a certain course. They listed a series of murderers [using GenAI]. (…) And then used those images to draw a new design in their sketchbook.” (Intw. 16)
Another example besides GenAI-generated images comes from a developer who uses GenAIgenerated voice lines in game prototypes: “if you start using those AI voices yourself to quickly put something together, so prototyping, (…) I also use those AI voices myself sometimes. And that's okay, as long as it's not in the final result” (Intw. 8). The final part of this developer’s response is the main reason why the interviewed developers view GenAI tools are more useful in early development stages: many of them report on ethical concerns of just copy-pasting GenAI-generated content into their games, or discuss concerns regarding copyright issues for their final gaming products (more on this in the next section).
Related to the notion that GenAI tools have a bigger role to play for programmers and executive roles in game development, respondents that had a neutral or positive attitude toward GenAI tools perceived such tools as more supportive, rather than creative, instruments. One respondent described it as this: “[Generative] AI helps. (…) You have to use it to help your creations, not to create” (Intw. 15). Interestingly, another developer mentioned that he predicts that the introduction of GenAI tools will be “more important for production-oriented changes than for changes in games” (Intw. 14), suggesting that GenAI will have a greater impact on how games are developed than on the games’ actual content.
Finally, only one respondent mentioned the potential strength of GenAI tools for indie game developers specifically: “[Generative] AI is still viewed very negatively at the moment, but I think it could be a good thing for indie companies to shoot above their ballpark” (Intw. 11).
This third theme represents the weaknesses and concerns associated with the integration of GenAI into game design. Five subthemes were found that discuss concerns about quality of GenAIgenerated output, risks regarding employment, copyright and other ethical concerns, and experiencing a lack of control with GenAI-generated content.
The biggest weakness that developers encountered was the lack of quality of current GenAI tools. These quality concerns exist for multiple roles within game studios—narrative designers, game artists, and even game programmers. One respondent mentioned that, when trying to develop more story-driven games with GenAI tools such as ChatGPT, “it’s relatively difficult to tame [generative] AI. Because they want to go very broad and don’t always have the right intentions when generating dialogue or storylines” (Intw. 17). GenAI tools are not just hard to ‘steer’ when trying to create storylines or dialogues. They also provide text that is easy to identify as being GenAI-generated, due to the perceived lack of storytelling quality: “While I understand that a game dev, who is not much concerned with story and has a nice idea for his gameplay, will ask ChatGPT: ‘What are nice environments or stories that I can tell?’ (…) But I personally find those answers so trite. I have tried it a few times, when I am in a kind of brainstorming kind of way… Give me ten funny wordplays (…) and they are all so bad. (…) It’s really just a search for banalities. A search for mediocrity.” (Intw. 9)
Another respondent mentioned that GenAI tools are “still a bit annoying to use, for example, for image generation. It’s just faster if we do it ourselves” (Intw. 11). Surprisingly, while GenAI was perceived as a positive innovation for coding and programming tasks, several respondents suggested that such tools can also have a negative impact on this type of labor: “for coding, it can sometimes help, but it can also sometimes be a very junior programmer and introduce sneaky bugs” (Intw. 17). Another developer connected this specifically to creating emotional experiences and losing control when trying to program such experiences using GenAI: “I rather stick to ‘human’ programming. Yes, it can go faster [for] programming. But I think if you really want to evoke specific emotions, you [can] lose some of that control” (Intw. 7).
Aside from quality concerns, the second most frequently mentioned risk among the interviewed developers was the potential threat of employment loss associated with the (mass) introduction of GenAI in game development processes. One respondent had a clear argument regarding this concern: “It shouldn't be at the expense of jobs (…). I think it's a shame that studios, mainly the larger ones, commit to [generative] AI models based on the work of their employees. So, suppose you are a good programmer or a good artist. You work for a large company. (…) And at the same time, there is an [generative] AI model behind the scenes that learns from that. And then a quarter of a year later, the company says ‘Yes guys, you did a good job, but I'm going to have to fire you.’ Because, in the meantime, we have created two [generative] AI models that can completely replace you.
That is a trend, for example. I am very afraid of that.” (Intw. 8)
Finally, several less frequently mentioned concerns relate to copyright and broader ethical risks, as well as experiencing a lack of control over GenAI-generated output. One respondent mentioned that GenAI-generated content cannot be included “into production, because then there are copyright issues. So, we’re not going to use that in production” (Intw. 1). Another developer shared his experience using GenAI tools to create mock-ups with AI-generated images. Although the client liked these exact mock-ups, the developer had to say that these exact images couldn’t be used as they were generated with GenAI, taking away employment from illustrators and graphic designers. He emphasized that these are ethical or deontological considerations each developer must take into account, given the lack of control over copyright and intellectual property. The final concern comes from a developer who mentioned that “people who are starting to use [and interact with generative] AI, they have them say and do certain things, giving them completely different meanings [from what developers intended]. That they start to become very inappropriate” (Intw. 7), suggesting that the introduction of GenAI tools, especially in character design, might lead to a lack of control from a developer’s standpoint over how players interact with such GenAI-powered NPCs.
The current paper’s objective was to provide a first, exploratory overview of how Belgian game
developers perceive and use generative AI in their daily game development. As prior research
focused only on student game developers [
In terms of general attitude toward GenAI’s integration in game development, a variety of
opinions existed, going from positive to downright critical attitudes. Most respondents, both
established developers and newcomers alike, were rather skeptical toward GenAI’s usefulness,
voicing clear concerns about the technology’s impact on quality, employment, and intellectual
property. These findings are not surprising, as they confirm earlier results among student developers
[
The identified strengths and opportunities of GenAI bare some similarities with prior work. For one,
developers showed a strong consensus toward the notion that GenAI tools are particularly useful for
technical tasks such as programming or coding over creative and artistic tasks [
Another interesting finding was that most of the current use of GenAI tools occurs during early
stages of game development, like gaining inspiration from ChatGPT [22] to give original names to
buildings or areas within the game, creating GenAI-generated artwork for mood boards, or use
GenAI-generated voices in early game prototypes. Using GenAI tools to jumpstart the development
process is an application of these tools that is also mentioned in prior work [
Regarding the weaknesses and concerns developers have about GenAI tools, the subthemes that were
identified in this study were fairly similar compared to the concerns mentioned in previous work [
Furthermore, career-related concerns were exclusively focused on the loss of employment by
fellow developers and artists, similar to prior research [
Finally, the main concern or weakness discerned about GenAI tools was their outputs’ lack of
quality, mentioned by artists, writers, and even programmers. While prior research showed that
mostly artists find the lack of quality disturbing [
There are two final findings of this study worth mentioning. First, the environmental impact of using GenAI was not a concern at all for the interviewed developers: sustainability was only briefly mentioned by one respondent. Second, concerns about developers’ lack of control over how players interact with NPCs powered by LLMs—potentially leading to inappropriate situations—have not, to my knowledge, been addressed in prior research. This last finding deserves additional investigation in future research.
Prior research by Vimpari et al. [
Common themes across these industries include a general concern regarding employment loss
due to the increasing integration of GenAI tools in game development workflows, which has also
been identified as a global concern [
Interestingly, while the Finnish and Polish game developers were generally more positive towards
GenAI tools—perceiving text-to-image-generating (TTIG) systems like DALL-E or Midjourney as
interesting and impressive [
The current study has several limitations to take into account when interpreting the findings. First, the choice to use a qualitative research design led to less generalizable findings. Despite respondents providing detailed and personalized insights into their attitudes toward and experiences with GenAI tools, the small sample size in this qualitative study is not necessarily representative of the Belgian game industry. As such, this study does not allow for generalizable conclusions, especially not for the global gaming industry, where varying contexts and levels of GenAI integration across regions may lead to different outcomes. Therefore, I would recommend future research on this topic to focus on conducting a large-scale survey of game developers, focusing on the Belgian industry or an a more international developer population. Such an endeavor could also compliment the current study’s bottom-up approach by building on existing theoretical frameworks, such as the Unified Theory of Acceptance and Use of Technology (UTAUT)[36]—a framework previously used to study the adoption of AI tools in other contexts [37] and which includes factors like performance and effort expectations, and social influences—to broaden explanatory factors regarding developers’ willingness to integrate GenAI in game development.
A second limitation relates to the exploratory nature of this study. Being part of a larger project on emotional or eudaimonic player experiences, examining the perceptions and usage of GenAI tools in developers’ practices was not the main focus of the interviews. This could explain why developers, for instance, did not mention the use of specific GenAI tools besides ChatGPT [22] as frequently as in previous work on this topic: we simply did not ask them to reflect that deeply on this topic. Similarly, the interview data lacked any clear findings about the use of GenAI for audio design or game marketing. While this may reflect a genuine absence of the use of such tools for these purposes, it could also be a result of the study’s exploratory setup. Building on this and previous research, future studies should therefore start with a clear focus on finding the strengths, weaknesses, opportunities and threats or risks accompanying GenAI integration in game development, for example, through a practice-based perspective like a SWOT analysis.
Additionally, future research should examine how GenAI tools, which increasingly offer automation benefits, affect not only the creation of artistic games— understood as games with experimental or unconventional narratives, mechanics, and aesthetics [38] that offer meaningful societal critique and novel storytelling [39]— but also how GenAI-powered game design affects players’ experiences, such as eudaimonia [32]. This research would address the existing tension between GenAI's capacity for development automation and the potential risk of diminishing the originality of and creative expression in artistic games. In doing so, such an approach can provide valuable insights for game developers, emphasizing a critical and careful approach to leverage its potential for economic competitiveness while safeguarding creative and artistic expression.
An important implication of the current study is that some identified weaknesses and concerns
could be decreased when game developers are properly trained to work with GenAI tools in the
context of designing games [19]. While the introduction of GenAI for game development has been
met with initial fears of employment reduction [
This exploratory qualitative study provides early insights into how generative AI affects game development in the Belgian gaming industry, revealing how developers perceive GenAI’s advantages and pitfalls, and how they currently use it in their development practices. Regarding general attitude, some developers expressed cautious optimism about GenAI’s potential, while most remained skeptical of its current practical value. GenAI’s strengths were seen in coding and other technical tasks rather than in creative work. Its opportunities lie mostly in early development stages—for instance, for generating names of in-game objects or creating images for concept art and mood boards. Quality concerns surfaced most often: Belgian developers criticized GenAI for mediocre storytelling, inefficient image generation, and buggy coding outputs. Worries about job displacement were also persistent, whereas ethical issues like copyright and IP infringement were mentioned less frequently.
These findings are relevant for several actors, including game developers, educators, policymakers, and researchers. They can inform discussions within the game developer community, advise policymakers in providing much needed legislation surrounding the boundaries of GenAI (and, for instance, clarifying copyright issues [19]), and may boost future research that contributes to a better understanding of how GenAI can be integrated in a supportive rather than harmful way in game development.
This interview study was part of a larger student research project of 3rd Bachelor students in Communication Sciences at the University of Antwerp, Belgium. As this was a student project and thus, part of a teaching assignment, no external funding supported this work. The author would like to acknowledge and thank students Laura Fransen, Wout Haagdorens, Yannis Ghyselen, and Tenzin Willaert for their help in setting up the research design, developing the semi-structured interview questionnaire, contacting the respondents, and for conducting and transcribing the interviews.
During the preparation of this paper, the author used GPT-3.5 and Gemini 2.5 Flash for grammar and spelling check, never to generate new sentences, arguments, or academic sources. After using these tools, the author critically reviewed and edited the generated content as needed, taking full responsibility for the paper’s final content. in: J. Nichols (Ed.), Proceedings of the ACM on Human-Computer Interaction, ACM, New York, NY, 2023, pp. 131–164. doi:10.1145/361102. [7] The AI Digital Quill, Beyond traditional play: The transformative impact of generative AI in gaming, February 5, 2024, Medium. URL: https://medium.com/@DigitalQuill.ai/comparativeanalysis-traditional-ai-generative-ai-in-the-game-industry-39144dbe0b67. [8] B. Marr, The role of generative AI in video game development, April 18, 2024, Forbes. URL: https://www.forbes.com/sites/bernardmarr/2024/04/18/the-role-of-generative-ai-in-videogame-development/. [9] M. Müller-Brockhausen, G. Barbero, M. Preuss, Chatter generation through language models, in: C. Harteveld, J. Liu, J. Togelius, G. Wallner, L. Chukoskie (Eds.), 2023 IEEE Conference on Games (CoG), IEEE, Boston, MA, 2023, pp. 1-6, doi:10.1109/CoG57401.2023.10333244. [10] Team Miaozi, Cygnus Enterprises, 2023. URL: https://steamcommunity.com/app/1963520. [11] R. Barth, The convergence of AI and creativity: Introducing Ghostwriter, March 21, 2023, Ubisoft. URL : https://news.ubisoft.com/en-au/article/7Cm07zbBGy4Xml6WgYi25d/theconvergence-of-ai-and-creativity-introducing-ghostwriter. [12] Midjourney Inc., Midjourney, 2023. URL: https://www.midjourney.com/home. [13] OpenAI, GPT Image 1, 2025. URL: https://poe.com/GPT-Image-1. [14] Tabnine, Tabnine, 2018. URL: https://www.tabnine.com/. [15] GitHub, OpenAI, GitHub Copilot, 2021. URL: https://github.com/features/copilot. [16] Video Games Europe, Call for contributions: Competition in ‘virtual worlds’ and generative artificial intelligence, July 12, 2024. URL: https://www.videogameseurope.eu/publication/callfor-contributions-competition-in-virtual-worlds-and-generative-artificial-intelligence/. [17] K. Monahan, G. Burlacu, From burnout to balance: AI-enhanced work models, July 23, 2024,
Upwork. URL: https://www.upwork.com/research/ai-enhanced-work-models [18] J. Becker, N. Rush, E. Barnes, D. Rein, Measuring the impact of early-2025 AI on experienced open-source developer productivity, arXiv preprint, July, 2025. doi:10.48550/arXiv.2507.09089 [19] L. Sikorski, J. Matulewski, M. Czerwonka, On the attitudes of gamedev industry artists toward genAI: Preliminary results, in: C. Riemenschneider, Y. Sullivan, M. Dinger, M. Bantan, N. Roberts (Eds.), Proceedings of the 2025 Computers and People Research Conference, SIGMISCPR ’25, ACM, New York, NY, 2025, pp. 1–6. doi:10.1145/3716489.3728432. [20] H.-K. Ko, G. Park, H. Jeon, J. Jo, J. Kim, J. Seo, Large-scale text-to-image generation models for visual artists’ creative work, in: Proceedings of the 28th International Conference on Intelligent User Interfaces, IUI ’23, ACM, New York, NY, 2023, pp. 919–933. doi:10.1145/3581641.3584078. [21] Adobe, Adobe Firefly, 2023. URL:https://www.adobe.com/products/firefly.html. [22] OpenAI, ChatGPT, 2022. URL: https://chatgpt.com/. [23] A. Christofferson, A. James, T. Rowland, I. Rey, How will generative AI change the video game industry?, September 2023, Bain & Company. URL: https://www.bain.com/insights/how-willgenerative-ai-change-the-video-game-industry/. [24] E. Strubell, A. Ganesh, A. McCallum, Energy and policy considerations for deep learning in NLP, in: A. Korhonen, D. Traum, L. Màrquez (Eds.), Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, ACL, Florence, Italy, 2019, pp. 3645–3650. doi:10.18653/v1/P19-1355. [25] European Video Games Society, Greening the video games industry: Winning solutions for the environment, Event report, Cologne, Germany, 2022. URL: https://digitalstrategy.ec.europa.eu/en/library/greening-video-games-industry-winning-solutionsenvironment. [26] S. Werning, Generative AI and the technological imaginary of game design, in: F. Lesage, M.
Terren (Eds.), Creative Tools and the Softwarization of Cultural Production, Springer, London, UK, 2024, pp. 67–90. doi:10.1007/978-3-031-45693-0. [27] V. Acovino, C. Intagliata, ‘Games made by soulless machines’: Tech sparks debate over AI stories in video games, March 15, 2024, NPR. URL: https://www.npr.org/2024/03/15/1238111971/video-games-ai-artificial-intelligence-nvidia. [28] J. Reed, Gen AI chatbots are starting to remember you. Should you let them?, April 25, 2025, CNET. URL: https://www.cnet.com/tech/services-and-software/gen-ai-chatbots-are-startingto-remember-you-should-you-let-them/. [29] BelgianGames, 2016. URL: https://www.belgiangames.be/. [30] A. Cote, J. G. Raz, In-depth interviews for games research, in: P. Lankoski, S. Björk (Eds.), Game Research Methods: An Overview, ETC Press, Pittsburgh, PA, 2015, pp. 93–116. doi:10.1184/R1/6686765.v1. [31] D. Possler, N. D. Bowman, R. Daneels, Explaining the formation of eudaimonic gaming experiences: A theoretical overview and systemization based on interactivity and game elements, Front. Commun. 8 (2023). doi:10.3389/fcomm.2023.1215960. [32] R. Daneels, N. D. Bowman, D. Possler, E. D. Mekler, The ‘eudaimonic experience’: A scoping review of the concept in digital games research, Media Commun. 9 (2021) 178–190. doi:10.17645/mac.v9i2.3824. [33] Lumivero, NVivo 1.7.2, 2024. URL: https://lumivero.com/products/nvivo/. [34] V. Braun, V. Clarke, Using thematic analysis in psychology, Qual. Res. Psychol. 3 (2006) 77– 101. doi:10.1191/1478088706qp063oa. [35] E. Gach, Report: Laid-off staff at Candy Crush maker say they’ve been training their AI replacements, July 15, 2025, Kotaku. URL: https://kotaku.com/candy-crush-ai-king-microsoftlayoffs-xbox-farm-saga-1851786283 [36] V. Venkatesh, M. G. Morris, G. B. Davis, F. D. Davis, User acceptance of information technology: Toward a unified view, MIS Quarterly 27 (2003) 425–478. doi:10.2307/30036540. [37] V. Venkatesh, Adoption and use of AI tools: A research agenda grounded in UTAUT, Ann.
Oper. Res. 308 (2022) 641–652. doi:10.1007/s10479-020-03918-9. [38] J. Juul, Handmade pixels: Independent video games and the quest for authenticity, MIT Press,
Cambridge, MA, 2019. [39] T. Holmes, Arcade classics spawn art? Current trends in the art game genre, in: Proceedings of the 5th International Digital Arts and Culture conference, RMIT University, Melbourne, Australia, 2003, pp. 46–52. URL: https://www.academia.edu/2998560/Arcade_Classics_Spawn_Art_Current_Trends_in_the_Art _Game_Genre. [40] V. Kovanovic, R. Marrone, MIT researchers say using ChatGPT can rot your brain. The truth is a little more complicated, June 23, 2025, The Conversation. URL: https://theconversation.com/mit-researchers-say-using-chatgpt-can-rot-your-brain-the-truthis-a-little-more-complicated-259450
Table 1 Overview of Sample Characteristics Interview Gender
Intw. 7 Intw. 8
45 37 29 32 49