What place will AI have in game development?

This article is part of AI Week.

The question of what place AI could and should hold in the games industry is one of the hottest and most divisive topics at the moment. And judging by the posts we often see on social media, the world can be neatly divided into AI evangelists and AI haters.

But of course, away from the polarising pile-ons seen in the more argumentative areas of the internet, the reality is somewhat different. GamesIndustry.biz asked a wide range of company heads about what place they throught AI would have in game development, and the answers were varied, but far more nuanced than the typical LinkedIn hottake.

Some are keen to explore the possibilities of AI, while others are vehemently against it, but the general consensus falls somewhere in between, with many seeing uses for AI somewhere in the development pipeline – even if they can’t fully embrace all its facets.

Tim Bender (CEO and Founder, Hooded Horse)

“I think GenAI art will be kept out of the best games from genres and developers where quality and craft are valued by players,” says Tim Bender, head of Manor Lords publisher Hooded Horse, who recently said his firm wouldn’t work with studios that use generative AI assets.

“Meanwhile, AI for productivity tasks or coding assistance – that is, for organising data or tools like Copilot – is going to see wider adoption. However, this will be in the nature of modest supplementary support and will not replace the central role of skilled programmers. Great games will still be made by great teams, and a lot of wild predictions going around lately will not come to pass – the recent stock price drop of game companies when Google’s Project Genie was shown is just silly.”

Jörg Tittel (Creative Director, RapidEyeMovers)

“AI will have all the places and no place at all in game development,” says Jörg Tittel, who was behind 2023’s C-Smash VRS. “Like with any tool, it’s our choice if, how, and where we use it. I personally work mostly with teams that are small, multitalented, and hands-on enough – we don’t have to use any AI in development.”

“For the next game, I did recently use AI to get a new concept across, but we quickly decided to scrap it. Doing my usual stick drawings would have been far more effective and suggestive. I might have felt more of a need to defend the idea.”

“It’s undeniable that any AI ‘art’ has inherently less value (carbon footprint aside). AI ‘art’ will only ever be based on what’s been made before, which goes counter to every true artist’s quest for authenticity and originality. Creation has to remain in the hands of the artist, otherwise you’re making (or not making) something hollow. I therefore draw the line at using AI anywhere near the ‘intent’ level, but see its value at the ‘content’ level.”

“Michelangelo did not paint the whole Sistine Chapel by himself, but employed a whole team of young artists to climb up ladders and risk their lives to copy his small ‘template’ painting and designs onto the enormous walls and ceiling. Video games can sometimes grow a million times larger in scope than Michelangelo’s masterpiece.”

“It’s undeniable that any AI ‘art’ has inherently less value”

Jörg Tittel

“I understand why people want to be employed, but I don’t see why skilled artists should be working under crunch conditions in order to texture a hundred photorealistic bins for the next open world game – or work stupid hours to make another cash-in HD remaster. AI can and will do that. It’s not just inevitable, it’s already happening. And perhaps that’s a good thing.”

“Fundamentally, we should get at each other’s throats less and let the informed ‘consumer’ decide. I, for one, will always opt for games made by humans and generated by original, organic thought.”

Susan Cummings (Co-Founder, 10six Games)

“Let’s be honest: most studios are looking at AI to shave zeroes off their budget,” says Susan Cummings, whose studio is behind the upcoming You vs Zombies. “They want cheaper assets, and faster code all to reduce dev timelines, but these experiments so far have come at a cost, replacing creativity with terrible results. While most of the industry is obsessed with efficiency, we are obsessed with magic.“

Cummings says the focus at 10six is to use new technologies to “create games that simply weren’t possible before,” adding that if “AI doesn’t make the game fundamentally more fun, we will find something else to make.”

“We believe we’ve found that magic. We’ve layered AI on top of our human-designed systems, not to replace the designer or artists, they are integral to this; but to empower the player. We can now let a player describe anyone, their loves, their hates, their look, and we’ve taught the AI to then refine that choice, and become a playable caricature in seconds.”

“It’s all about agency. We use AI as the difference between picking from a list of pre-made skins and genuinely injecting yourself into the world.”

Alexandre Amancio (SVP of World-Building and IP Strategy, FunPlus)

“Firstly, we need to stop treating AI like magic or ideology – it’s neither,” says Alexandre Amancio, who was the creative director of Assassin’s Creed: Revelations and Assassin’s Creed: Unity at Ubisoft, and recently talked about the problems facing AAA development.

“At its core, AI is a statistical tool. It produces the most probable output based patterns it has learned from training data. That makes it incredibly powerful (and fast) for execution, but fundamentally different from human creativity.”

“Used correctly, AI is a game-changer. It removes friction, accelerates iteration, and allows small, focused teams to achieve things that previously required industrial-scale production… and the operating budget of a small city. In many ways, it’s one of the most democratizing forces game development has seen since free game engines.”

“But there’s a real danger if we start outsourcing judgment itself. AI can generate novel combinations, but it lacks the elements that drive true creative breakthroughs: intent, taste, and most importantly, lived perspective.”

“Culture moves forward because of outliers: strange ideas, strong voices, and creative risks that don’t initially look statistically correct. The creators that redefine the medium rarely follow probability curves. They break them.”

“So the future of AI in game development isn’t predetermined. It’s a choice. One path leads to an industry flooded with competent but indistinguishable experiences: efficient, technically impressive, and creatively hollow. The other uses AI as a catalyst to empower human vision, enabling more singular voices, more experimentation, and more unexpected ideas to reach players.”

“We need to decide whether AI serves creativity, or whether creativity becomes subordinate to efficiency”

Alexandre Amancio

“AI isn’t going away. Ignoring it is naïve. But blindly embracing it without protecting human creativity is outright dangerous.”

“AI doesn’t decide what matters. Humans do. I think the real challenge isn’t technological; it’s cultural. We need to decide whether AI serves creativity, or whether creativity becomes subordinate to efficiency.”

“The industry needs to come together, in a way it never has before, and have difficult, meaningful, open discussions on the matter.”

Hendrik Lesser (Founder and CEO, Remote Control Productions)

“AI isn’t automation – it’s amplification,” says Hendrik Lesser, who is president of the European Game Developers Foundation in addition to being the founder of Lesser Evil and Remote Control Productions. “We’ve moved past ‘is AI good or bad?’ The real question: how do we stay in the driver’s seat?”

“Think generative coding, management AI, and tools. Magic happens when you connect all three. But you must remain the director. Make the calls. Keep the human touch.”

“The premium will always be vision, culture, stories that resonate because someone cared. That’s not anti-AI. That’s pro-creator with a director’s mindset.”

Tracey McGarrigan (CEO and Founder, Ansible Comms)

“The debate around AI in games right now gets framed as a battle for the industry’s soul, but I see it as a generational transition,” says Tracey McGarrigan, who acts as a strategic marketing advisor to a number of firms, including 10six Games. “While many are rightly protective of the craft, we can’t ignore the fact that traditional GTM [go-to-market] models alongside the increasingly redundant definition of a ‘gamer’ are shifting.”

“My observation is that the games industry is avoiding direct eye contact with AI while signals from the broader world are brightly shining. At BETT this year, we saw that 85% of EdTech on show is now AI-integrated. This next generation is being educated to see AI as a collaborative tool that responds to their intent. If Millennials have ‘mechanical literacy’ (knowing which buttons to press), Gen A and Gen B will have ‘intent-based literacy’ (knowing how to prompt or direct a system to achieve a result).”

“The biggest hurdle for game creative leads is that this isn’t just about the tech, it’s now about commercial survival. If you are only using AI to shave your budgets or eliminate the need for people on your team, that is a race to the bottom. AI should be seen as a tool to augment talent, not replace it.”

“AI is critical here because it can strip away some of the heavy technical weight and investment that can bury great ideas, and sets us up to ‘find the fun’ or test market-fit far quicker than traditional cycles allow. Too often I hear devs fed the hollow advice to ‘just make the game you want to make’ without a commercial plan. That’s like telling a golfer to ‘just get a hole-in-one’ and calling it a business model. It’s not a strategy; it’s a gamble.”

“It’s about creators finding new ways to design fun that wasn’t possible a couple of years ago”

Tracey McGarrigan

“But it’s also not about asking AI to make the game for us, nor is it about procedural slop; it’s about creators finding new ways to design fun that wasn’t possible a couple of years ago. It’s about putting tools in the hands of those willing to reach an audience many aren’t catering for yet. Whether it’s then delivering hyper-personalization and player agency in runtime, or opening up instant, no-code world-building where we can test game ideas live in a browser, AI supports us in serving audiences that a pocket of the traditional industry is trying to convince itself doesn’t exist.”

“I don’t see AI as a threat to the craft. I’m witnessing every day how smart, crazy talented developers are using it to build sustainable game businesses for a generation that values creative agency over passive consumption. When people tell me this isn’t what players want, I push back. It might be that these new games don’t appeal to everyone, nor should they. That should also be okay, there’s room for everyone, yet I am so disappointed at how dismissive some people are towards developers experimenting with new ideas and tools. I genuinely love and respect all game makers, but right now, I’m backing those using AI to find the fun for an audience that is already trying out new ways of playing.”

Joe Harford (Founder and CEO, Airship)

“When people say ‘AI in game dev’ they jump straight to generative art, but it’s so much broader than that,” says Joe Harford, head of the art production company Airship. “It touches every part of the creative process: game design, level design, writing, audio, QA, tooling. And honestly? We’ve been chasing faster, more efficient ways to make games for decades. That’s not new. Funders want shorter dev cycles, studios want to ship more efficiently, and every major tooling leap we’ve had, from better engines to procedural generation to machine learning pipelines, is in some way artificially intelligent. It’s a spectrum, not a switch.”

“My view is that, where morally and legally usable, it’s fine, more than fine, so long as enough of the process is guided and steered by humans who have a genuine creative understanding of what they’re building.”

“Could AI build full games in the future? Probably. It can already generate video, text, and graphic content at scale. But it can also create slop, in the same way it creates slop images, slop writing, slop videos that fool people for five minutes. Even when it gets genuinely good at game creation, and it will, we still value the knowledge that something came from a human. That human experience in game development is what anchors the developer to the player. It’s the bond. It’s the connection.”

“There have always been games you could call dev slop. No care for the player, just chasing the next trend or the next revenue line. Good games have always stood out through genuinely unique experiences, good journalism, and solid marketing. It takes a community. A collective. And I think we’d be better served spending less energy discouraging AI tools which are here to stay and more energy encouraging and promoting crafted human experiences.”

“My view is that, where morally and legally usable, it’s fine, more than fine, so long as enough of the process is guided and steered by humans”

Joe Harford

“Team sizes will shrink. Tools will evolve. Studios will lean more on external partners and outsourced teams who can flex and adapt, teams who are already adapting to these new tools and workflows. That’s the reality. But at its heart, game development is an art form and an entertainment form, and there’s room for all types of games for all types of players.”

“The legitimacy question is where it gets important though. There’s a massive difference between procedurally generated dungeons in an ARPG, which players love, and hundreds of AI-generated slop games pumped out to farm micro-revenue across a sea of garbage. We’ve already seen what that looks like with AI children’s content on YouTube, and it’s grim.”

“We all have a duty to be more discerning in what we play, what we let our kids play, and what we recommend. The truly unique experiences that humans make will always stand out, but only if we give them the room to. That’s where I think AI sits in the future of game development. Not as the replacement, but as the thing that makes the human craft even more visible by contrast.”

Shalin Shodhan (Director, Masala Games)

“It breaks my heart every time I see a billboard in my city with Ghibli-style AI slop used as art work,” says Shalin Shodhan, whose Ahmedabad-based company released Detective Dotson last year. “Studio Ghibli didn’t want it, they didn’t get compensated for it, and the AI company got viral growth out of it. It’s not right. I have nothing against Stable Diffusion or StyleGAN: they are brilliant technical advances, but the results are ugly/scary/soulless, and the sources have to be compensated.”

“Somehow, I don’t have the same outrage/heartbreak with LLM-generated text and code. I have fully embraced coding agents.”

“Once you get the hang of communicating well with your agent, it truly makes the iterative game development process fun. The amount of time I spend testing my game has gone up manifold, since I’m afforded so many more iterations. It’s a bit of a golden age for designers and product thinkers to be building prototypes and then, with the right processes, even turning them into full-fledged products.”

“In pre-AI gamedev we used game engines, and the standard ones were not great at deploying to the web. So we ended up distributing on closed platforms that take 30% in fees. But with AI, you don’t need to be wed to an engine – just make your own! Since the web is the most openly available training set available, AI is best positioned to make web games. I’m most excited about this. I hope AI-based gamedev brings a revolution in web games. Just awesome, wild experiences that are pure fun, available to everyone everywhere.”

Brandon Sheffield (Creative Director, Necrosoft Games)

“I think that until the bubble bursts, AI will be used by quite a few game companies, and perhaps even beyond the bubble,” says Brandon Sheffield of Necrosoft Games, which released the RPG Demonschool last year. “Look, I am not a fool – I understand that AI has some uses. Primary among those being the devaluing of work, the cheapening of labour, the laying off of creatives, and the generation of busywork for humans to fix whatever AI generates.”

“It’s also useful for burning through water, raising the prices of RAM and chipsets globally, and poisoning anyone near the data centres that run them. So there are a lot of use cases for AI. From spurring on the death of the creative endeavour to dissolving our economy, AI is absolutely gangbusters at what it does. We don’t use it in our studio though.”

John Clark (independent consultant)

“AI will continue to be a polarising subject for some time to come,” says John Clark, who spent 12 years at Sega Europe. He subsequently worked for Tencent, and later became CEO of Curve Games, but nowadays he works as an independent consultant, and is an advisor for the AI start-up Infinity Fiction.

“AI challenges values around creativity and job security, brings legal insecurity, and risks brand identity. However, AI can improve iteration speed, improve live ops, expand what small teams can achieve, and help developers solve problems faster. Making games can be more affordable.”

“I think about AI in games in two ways. The first is about the operational aspects of game development in terms of removing friction from development, driving operational excellence, improving efficiencies, and improving ROI. QA, localisation, bug triaging, live ops support, customer support, and so on should benefit.”

“Guard rails are required, and IP protection, licensing, and copyright are all vital”

John Clark

“Where AI is more challenging is with core development, where it is seen to replace development craft and creativity. I’m excited to see how AI can ‘scale’ human creativity, how games can be made more reactive rather than scripted, and how simulated worlds can be enhanced. However, guard rails are required, and IP protection, licensing, and copyright are all vital. This is more sensitive and dependent on the philosophy of the studio.”

“I expect a period of misalignment between those who approach AI from a cost, scalability, and predictability perspective and those who primarily see the threat to quality, creativity, originality and reputation.”

“I think both are right. Bringing the costs down, being quicker to market, and building predictable models is vital. However, the amazing creativity that attracts us all to video games as players and a business is what drives the entire industry.”

Harish Chengaiah (Founder, Outlier Games)

“Any task that’s menial, repetitive, and time intensive is a good domain for AI usage,” says Harish Chengaiah of Outlier Games in Chennai, India, who has recently been working with the government of Tamil Nadu on policies affecting the country’s games industry. “We already have procedural generation, and going forward, I see the animation pipeline benefiting the most from AI usage: we already have an example of this with the animation tool Cascadeur.”

“But AI would be a very bad fit for any task that operates in the domain of emotional expression, as emotions gain meaning from the uncertainty and vulnerability of humans’ limited but embodied lifespans. This gives humans a very different view of all the emotions felt and how we express them. AI cannot truly grasp this due to its permanent, invulnerable, unconstrained, and non-existential nature, and any attempt to do so will always feel uncanny or not measure up to something that humans do. For this reason, I anticipate that once the hype around AI settles, the market will realise that creative and expressive jobs will anchor employment in the coming era.”

Mario Wynands (CEO, PikPok)

“AI already has a foot in the door, as I can tell there is already quite a lot of adoption across the games industry, mostly only disclosed in hushed tones in the dark corners of industry cocktail mixers, given the controversial nature of the technology,” says Mario Wynands, head of the New Zealand-based games developer and publisher PikPok.

“Many developers and publishers are already chasing the promise of cost savings, faster time to market, and/or higher productivity that AI ‘promises’, although most are realising only incremental gains at best, and even wasting time and resources at worst.”

“The challenges to achieving more complete adoption and benefits are many, with very real legal, ethical, morale, and environmental issues, negative consumer sentiment, and, fundamentally, the utility, quality, and ‘creativity’ of AI tools often falling short on both requirements and expectations. The games industry just isn’t positioned for high utilization of AI to lead to critical and commercial success, as evidenced by the lack of a complete, good, ‘AI created’ game in the top selling charts or award nominations.”

“But unlike blockchain and NFT gaming – which had a similar arrival into the games industry, but have mostly since been expunged from it – I do think AI is probably here to stay. ‘AI’ is multifacted, with many potential applications, massive ongoing investment, and a rapid pace of change. The legal issues will be resolved for better or worse over time, some ethical solutions will arise, and, perhaps, energy efficiencies will be found.”

“Keeping AI out of game development will ultimately be like King Canute trying to hold back the sea”

Mario Wynands

“In all likelihood, pockets of consumer apathy will set in if an ‘AI forward’ iteration of a fan favourite franchise or two turns out to be ‘good’, and/or if AI capability can deliver a compelling experience that otherwise wouldn’t be possible. Existing AI tools will improve, new tools, models, and paradigms will arrive, and AI integrations will continue to creep deeper into industry-standard tools, whether you want them or not. Many venture capital firms and publishers will probably exert pressure on developers to leverage AI to better spend their supplied funding. Keeping AI out of game development will ultimately be like King Canute trying to hold back the sea.”

“That isn’t to say AI will end up in every corner of the games industry. There will always be a place for the auteur or small team to make a completely hand-crafted game experience, and that might even attract a premium in the way artisanal ceramics do over mass-produced knick knacks. But there will probably come a time for most studios when the technology is mature and the market is such that the benefits become not only desirable, but also actually necessary in order to remain competitive and commercially viable. And at that point, we’ll all have to make a choice.”

“The real challenge here for the games industry – and indeed society at large as it wrestles with the implications of AI technology – is that we must continue to be human-centric in our approach. We should be considered and deliberate. We shouldn’t defer decision making or creativity to AI. We need to be conscious of the costs as well as the opportunities. If we are going to entertain AI, we must make unlocking human potential the goal, not human replacement.”

This article was amended to restore some cut portions of Harish Chengaiah’s statement.